Neuralink
旗舰级侵入式 BCI 平台已经拿到真实临床动能,但在约 $9B-$9.6B 的偏高价格锚点下,公开经济性、报销准备度和长期支持义务仍披露不足。
Neuralink 是侵入式 BCI 里的旗舰平台,临床和融资动能都可信;但公开经济性缺失,报销与长期支持依赖仍未解决,~$9B-$9.6B 的估值锚也偏紧,因此仍只能继续研究。
封面要素
公司概况
Neuralink 是一家 2016 年创立的私营神经技术公司,做的是侵入式脑机接口平台,而不是纯软件 AI 或通用机器人业务。公开证据支持把加州 Fremont 视为最稳妥的总部信号,也支持将公司归为后期私营临床开发企业:它已从 2023 年获得 FDA 首次人体研究批准,推进到 2024 年首例患者验证、国际试验扩张,并在 2025 年完成 $650M Series E,但仍未公开披露商业收入。当前产品栈把 N1 植入物、机器人植入、解码软件和医院主导的临床流程组合起来,服务于瘫痪患者的设备控制,同时把路线图延伸到辅助机械臂控制和视觉恢复。
- 创始人
- Elon Musk, Dongjin "DJ" Seo
- 总部
- Fremont, California, United States
- 产品
- Neuralink 当前平台是一套围绕 N1 脑植入物搭建的托管式植入服务流程,包含机器人插入系统、设备端信号处理、无线解码软件,以及医院主导的支持,帮助瘫痪人群使用电脑、手机和相关外部设备。路线图还包括用于辅助机械臂控制的 CONVOY,以及未来视觉假体项目 Blindsight。
- 客户
- 当前终端用户是重度瘫痪成人,尤其是因脊髓损伤或 ALS 导致四肢瘫痪的人群;现实买方和运营伙伴则是具备试验能力的医院、神经外科团队和研究运营方,而不是自助式消费者。
- 商业模式
- 公开证据支持一种由医院和支付方中介的植入模式:Neuralink 最终通过植入硬件、手术、解码器支持和随访服务变现,交付路径经过专门临床站点,而不是消费订阅或大范围直销。
- 阶段
- Late-stage private clinical development
- 融资情况
- 最新披露融资是 2025-06-02 宣布的 $650M Series E,主要既有投资方参与。围绕该轮的公开报道指向约 $9B 投前和约 $9.6B 投后,但第三方来源对累计融资总额仍给出不一致口径。
执行摘要
主要优势
- Neuralink 拼出了业内最强的公开全栈侵入式 BCI 叙事:植入硬件、机器人植入、解码软件和多适应症路线图同时推进,而不是只讲单一组件。
- 临床证据已不只是概念:2024 年首位患者开始使用,公开更新到 2026 年 1 月已覆盖 21 名参与者,站点网络也从美国扩到更多国际项目。
- 公司在这一品类里的融资能力仍然突出,June 2025 $650M Series E 和顶级投资人阵容就是明确信号。
- CONVOY 机械臂工作和 Blindsight 的 FDA Breakthrough Device Designation,已经显示平台价值不止于光标控制。
主要风险
- 公司仍处临床阶段,收入、定价、毛利率、烧钱速度和现金跑道都没有公开披露,暂时无法按经营基本面测算。
- 商业化还要等 FDA 上市许可、编码和报销架构,以及可持续的试验后支持义务;公开记录尚未证明这些问题已经跑通。
- 设备可靠性和连续性风险仍然真实存在:Neuralink 披露首位参与者出现电极线回缩,公开证据也还缺少翻修率、不良事件和取出储备细节。
- 治理和信任包袱仍在:动物福利审查、劳动指控、高管职责映射不透明,以及部分负面事项缺少文件层面的闭环,都会拉大技术内在价值的折价。
- 按大约 2025/2026 的价格锚看,已披露装机基数很小,公开商业证据又缺席,估值显得偏高。
未决问题
- 公开收入桥仍不存在:定价、报销回款、收入结构、毛利率、烧钱、债务和现金跑道均未披露。
- 累计融资额以及完整股权结构表 / 清算优先权结构,在留存公开证据中还无法对齐。
- 公开证据没有量化翻修手术、严重不良事件、停用、队列留存,或正式的取出 / 连续性储备。
- 具体报销计划——编码路径、CMS 沟通细节和商业保险进展——仍未披露到投资人可测算的程度。
- 治理能见度仍不完整,因为已审阅的官方材料没有公布完整董事会名单,也没有清晰的高管治理图。
目录
01公司概况
1.1 身份、产品模式与当前阶段
Neuralink 是一家私营神经技术公司,正在开发可植入脑机接口,而不是纯软件 AI 业务或通用机器人平台。官网始终把公司定位在脑到设备控制上,第三方证据则把其当前总部锚定在加州 Fremont,并显示运营足迹覆盖加州和得州。当前产品故事仍处于临床阶段:公司通过 PRIME 及相关项目招募并植入患者,而不是销售一款广泛可得的商业设备。 公开证据同样支持把 Neuralink 定义为后期临床开发公司,而不是收入阶段公司。公司已从 2023 年 FDA 试验批准推进到 2024 年首例患者证据和 2025 年大额 Series E 融资,但本报告审阅的来源没有披露收入、ARR 或客户数。Sacra 对商业模式的表述——通过医院销售、附带手术和支持的垂直整合医疗器械系统——方向上合理,但它仍是目标变现路径,不是已经验证的运营模式。这个区别很重要:后续尽调章节应把临床执行和资本充足性,而不是既有商业规模,作为公司概况的基线。[CO001, CO002, CO004, CO005, CO006, CO007]
| 指标 | 数值 / 状态 | 日期 | 置信度 | 缺口 / 备注 |
|---|---|---|---|---|
| 成立时间 | 2016 | 2016 | 中 | 多个第三方来源相互印证,但抓取到的官方文本没有突出说明。 |
| 总部 | Fremont, California | 2026-06-08 | 中 | 基于 TipRanks 列表,而不是清晰的官方地址页面。 |
| 运营足迹 | California 和 Texas | 2022-2026 公开报道 | 中 | 公开证据支持设施 / 实验存在,但不是完整站点地图。 |
| 阶段 | 后期私营临床开发 | 2026-06-13 | 中 | 人体试验和融资公开;未披露商业指标。 |
| 最新融资 | 已宣布 $650M Series E | 2025-06-02 | 高 | 公司官方公告得到 CNBC 和 TechCrunch 印证。 |
| 估值信号 | $9.0B-$9.6B 头部区间 / $9B 投前报道 | 2025-05 至 2026 快照 | 中 | 已审阅官方材料未披露公司验证的投后估值。 |
| 累计融资 | 当前 | 低 | 第三方统计相互冲突,包括 Sacra 内部不一致的数据。 | |
| 员工数 | 681 名员工 | 2026-06-08 | 中 | TipRanks 的第三方估计。 |
| 收入 / ARR | 当前 | 低 | 已审阅来源未发现公开披露。 | |
| 客户数 | 当前 | 低 | 试验参与者有选择性披露,但未披露商业客户基础。 | |
| 人体试验状态 | PRIME 招募中;首位患者 2024 年植入;公司称截至 2025 年中已有五例植入 | 2023-2025 | 中 | 超出公司公布试验更新之外的规模仍有限。 |
Null 表示该指标未公开披露,或无法从可支持来源中调和。估值和累计融资行故意保留歧义,而不是强行给出单一数字。
[CO003, CO004, CO005, CO006, CO020, CO024]截至 2026-06-13,本轮报告对阶段、资本、人员和披露质量的压缩读数。
[CO020, CO024, CO028, CO038, CO039, CO040]1.2 创始人、领导层、治理与关键人物风险
公开记录支持 2016 年创立这一时间点,但与 Neuralink 的技术叙事相比,规范治理信息要薄得多。Yahoo Finance 转引的 Fortune 报道称,Elon Musk 与七名科学家和一名工程师共同启动公司,原始团队中只剩 Musk 和总裁 DJ Seo。TipRanks 目前列出 Jared Birchall 为 CEO、Musk 为联合创始人 / 联席 CEO、DJ Seo 为联合创始人 / 总裁 / COO、Shivon Zilis 为运营和特殊项目总监。这给出了可用的运营地图,但仍是间接证据,因为本次审阅的官方页面没有公布董事会名单,也没有清晰的高管团队页面。 头衔记录也不一致。Fortune/Yahoo 称 Birchall 在文件中被列为 CFO 等职务,TipRanks 则称他为 CEO。这不妨碍分析,但提出了治理尽调问题:外部人能推断谁重要,却无法清楚核实谁正式承担哪项高管授权。因此,关键人物依赖看起来高度集中在几个人身上:Musk 是品牌、资本磁铁和战略赞助人;DJ Seo 提供创始技术连续性;Birchall 则是与 Musk 相关的行政操盘者。公司仍为私营且披露较少,这种集中度应被视为风险放大器。[CO003, CO025, CO026, CO027, CO028, CO032]
| 人物 | 公开角色 / 证据 | 背景或职能覆盖 | 创始人连续性 / 治理信号 | 关键人物依赖 |
|---|---|---|---|---|
| Elon Musk | 联合创始人;TipRanks 也标注为联席 CEO | 品牌、融资磁铁、战略赞助者、Neuralink 公众面孔 | 已审阅全部公开记录中唯一持续可见的创始人 | 高——资本获取、招聘吸引力和公共叙事集中 |
| Dongjin “DJ” Seo | TipRanks 中为联合创始人、总裁兼 COO;Fortune/Yahoo 称其为仍在公司的创始人 | 贯穿设备、植入和产品转化的技术连续性 | 据称仍在公司的两名原始创始成员之一 | 高——创始技术连续性和项目可信度 |
| Jared Birchall | TipRanks 列为 CEO;Fortune/Yahoo 称文件中还有 CFO 等头衔 | 与 Musk 生态紧密绑定的行政运营者 | 头衔不一致本身就是治理尽调信号 | 中高——运营连续性存在,但正式授权不透明 |
| Shivon Zilis | TipRanks 称其负责运营和特别项目;Fortune/Yahoo 画像显示任期较长 | 跨职能运营者,连接内部执行与 Musk 网络协调 | 不是创始人,但作为私营公司运营者异常可见 | 中——支持执行,但角色边界未正式公布 |
覆盖范围有意限于公开可识别且实质影响公司方向的领导者。公开高管头衔证据不一致,尤其是 Birchall;本次运行未找到官方董事会披露。
[CO025, CO026, CO027, CO028]当前 Neuralink 故事里,产品野心、资本、监管方、运营者和患者证据如何互相连接。
[CO006, CO007, CO025, CO028, CO029, CO036]1.3 融资历史、估值信号与利益相关方地图
Neuralink 的资本故事在单轮融资公告层面清晰得多,在累计融资层面则模糊得多。最强锚点是 2025 年 6 月官方宣布的 $650 million Series E,CNBC 和 TechCrunch 均有佐证。公开报道还点名了一个可识别的投资方组合,包括 ARK Invest、Founders Fund、Sequoia、Thrive 和 Lightspeed。早期报道补充了 $280 million Series D 以及之后 $43 million 的追加批次,这有助于解释为什么第三方数据库显示不同的累计金额。 问题不是融资证据不足,而是缺少一本权威台账。TechCrunch 在 2025 年交割前引用了约 $9 billion 的投前估值;Sacra 目前既显示 $9.0 billion 的估值标记,又在叙事中称 2026 Series E 录入后新股融资约 $1.85 billion。这与同一页面上更保守的 $1.20 billion 融资页眉有实质差异。因此,本章把单轮事实视为可靠,但把累计融资和估值数字按区间处理。TipRanks 估计 681 名员工,也强化了这幅图景:一家资本充足的私营公司在收入、客户数或股权集中度披露更充分之前,已经先扩大员工规模。[CO020, CO021, CO022, CO023, CO024, CO033]
| 利益相关方 | 角色 | 控制或经济重要性 | 证据 | 尽调要求 |
|---|---|---|---|---|
| Elon Musk | 创始人 / 战略赞助者 | 主要公众身份,并可能对资本和叙事有决定性影响 | Fortune/Yahoo 创始人连续性;TipRanks 领导层列表 | 澄清正式董事席位、投票控制权和日常决策权。 |
| DJ Seo | 创始技术运营者 | 从原始创始团队延续到当前项目执行 | TipRanks 领导层列表;Verge 展示会角色;Fortune/Yahoo 连续性说明 | 澄清其对 R&D、临床和产品决策的权限范围。 |
| Jared Birchall | 与 Musk 关联的行政高管 | 可能是财务、法律或高管协调的控制点 | TipRanks CEO 标签与 Fortune/Yahoo 的 CFO 式描述冲突 | 解决正式头衔、职责范围和汇报线。 |
| Series E 投资财团 | 资本提供方 | 具名投资人为后期扩张提供资金和外部验证 | CNBC 和 TechCrunch 点名 ARK、Founders Fund、Sequoia、Thrive、Lightspeed | 要求股权结构表,检查按比例跟投权,并识别领投方治理权。 |
| FDA / IRB / 医院站点 | 临床闸门方 | 人体试验进展和任何商业化路径都依赖监管机构与临床站点配合 | CNBC 批准 / 招募报道,以及 PRIME 研究引用 | 要求协议状态、站点数量,以及任何暂停或修订。 |
| 试验参与者 | 使用证明利益相关方 | 早期患者结果是今天唯一公开产品牵引信号 | CNBC 首位患者、国际象棋演示,以及 TechCrunch 五例植入更新 | 要求队列规模、留存、不良事件报告和终点定义。 |
已审阅来源未公开披露经济所有权比例和董事会权利,因此该图谱强调影响路径,而非精确持股。
[CO020, CO021, CO022, CO025, CO026, CO028]1.4 里程碑、监管进展与反向信号
Neuralink 的时间线足够扎实,可以作为后续章节复用的基线。公司从 2022 年展示会式招募推进,到 2023 年 5 月获得 FDA 首次人体研究批准,2023 年 9 月启动 PRIME 招募,2024 年 1 月完成首例植入,2024 年 3 月公开首例患者验证,2024 年 5 月披露线体回缩问题。随后 TechCrunch 报道称,Blindsight 在 2024 年 9 月获得突破性设备里程碑;到 2025 年 6 月,公司正式完成 Series E,并声称已有五名重度瘫痪患者接受植入。 这种向前推进没有抹掉反向记录。CNBC 记录了危险材料运输调查;Reuters/CNBC 报道 FDA 曾因核心安全问题拒绝试验申请;Reuters/Inc 称动物实验室因令人反感的条件受到 FDA 批评。PCRM 走得更远,公布了与 UC Davis 相关的猴子实验图文指控。这些问题单独看都不能证明临床项目会失败,但合在一起说明,Neuralink 的进展来自一条充满监管和伦理争议的路径,而不是干净、低摩擦的执行。因此,下方里程碑时间线既是增长叙事,也是风险叙事。[CO008, CO009, CO010, CO011, CO012, CO013]
| 日期 | 事件 | 类型 | 金额 / 估值 / 状态 | 参与方 | 含义 |
|---|---|---|---|---|---|
| 2016 | Neuralink 成立 | 创立 | Musk 和创始科学家 / 工程师团队 | 后续全部时间线的起点。 | |
| 2022-11 | 展示会推动人体试验时间表和产品招聘 | 治理 | 人体试验被描述为约 6 个月后 | Musk、DJ Seo、招聘受众 | 获批前已释放紧迫感和产品化压力。 |
| 2023-02 | DOT 对被指不安全运输污染硬件的调查公开 | 反向 | 据报道有联邦调查 | CNBC、PCRM、DOT 发言人 | 引入非临床运营风险叙事。 |
| 2023-05 | FDA 批准首个人体临床研究 | 监管 | 首个人体研究获准推进 | FDA、Neuralink | 临床推进的关键解锁。 |
| 2023-09 | PRIME 研究开放招募 | 监管 | IRB 和医院站点批准后启动招募 | Neuralink、IRB、医院站点 | 公司从获批转向实际患者漏斗。 |
| 2023-11 | SEC 文件披露 Series D 追加融资 | 融资 | 在此前 $280M 分批融资基础上追加 $43M | Founders Fund 领投的投资人团 | 首例植入前延长资本跑道。 |
| 2024-01 | 首例人体植入披露 | 产品 | 患者恢复良好 | Neuralink、首位患者 | 首次证明 Neuralink 达成人体植入里程碑。 |
| 2024-03 | 首位患者公开演示光标控制和下棋 | 产品 | Noland Arbaugh 被展示使用植入物 | Neuralink、Arbaugh | 提供公开功能证据。 |
| 2024-05 | 导线回缩问题公开 | 反向 | 有效电极减少;已部署变通方案 | Neuralink、Arbaugh | 即便有了人体证明点,技术脆弱性仍然存在。 |
| 2024-09 | Blindsight 获得 FDA 突破性设备认定 | 监管 | 据报道获得突破性设备认定 | FDA、Neuralink | 产品叙事扩展到运动控制之外,但仍处临床前。 |
| 2025-06 | Series E 完成,公司称已有五人接受植入 | 融资 | $650M;公开报道中的约 $9B-$9.6B 估值区间 | Neuralink 及具名投资财团 | 把资本扩张与早期多患者信号结合在一起。 |
| 2023-06(后续报道) | FDA 动物实验室检查发现需整改的问题 | 反向 | 被敦促自愿整改 | FDA、Reuters/Inc、Neuralink 动物实验室 | 让监管和伦理审视留在正式时间线中。 |
这是本章唯一的正式时间线。日期采用最可支持的公开时间;若后续报道披露了更早事件,本行保留底层事件日期,并在含义列解释背景。
[CO003, CO009, CO010, CO011, CO012, CO013]从创立到 2025 年 Series E 轮以及多患者更新,几个带日期的关键拐点。
[CO003, CO009, CO010, CO011, CO012, CO013]1.5 图表
02市场分析
2.1 市场边界、替代品与相邻赛道
Neuralink 正确的市场边界不是“全部神经技术”,甚至也不是每一种脑机接口产品。Neuralink 当前公开提供的是一款研究性可植入 BCI,面向因脊髓损伤或 ALS 导致四肢瘫痪的成人,通过机器人辅助手术交付,目标是恢复高带宽电脑控制。因此,应纳入的支出包括植入物、手术流程、解码软件、临床编程,以及让系统能在医院到家庭路径中可用的持续支持。这个边界不能干净地覆盖消费级 EEG 头戴设备、游戏接口、泛 AI 神经技术叙事,甚至也不覆盖全部神经刺激。今天的现状替代品包括用于沟通的眼动追踪 AAC 和 Medicare 覆盖的语音生成设备,以及 DBS 等服务于其他神经学任务的成熟植入式神经刺激。外骨骼、中风康复 BCI 和未来运动恢复系统等相邻类别有战略意义,但应被视为扩张向量,而不是 Neuralink 已经打开的可服务市场。[CM001, CM002, CM003, CM004, CM005, CM006]
| 细分 / 类别 | 纳入支出 | 排除支出 | 买方 / 支付方 | 与 Neuralink 的相关性 |
|---|---|---|---|---|
| 可植入辅助型 BCI 核心 | 面向四肢瘫痪成人的植入物、机器人手术、解码软件、编程、随访,以及辅助计算机控制工作流 | 消费级 EEG 头显、健康神经科技、无关机器人 | 可开展试验的医院;当前为研究资助方 | Neuralink 近期核心市场边界 |
| 沟通恢复替代方案 | 眼动追踪 AAC、语音生成设备工作流、护理者支持的沟通技术 | 非沟通类消费电子 | 患者、护理者、言语团队、DME 覆盖 | 现状替代方案,会限制接受手术的意愿 |
| 植入式神经刺激邻近市场 | 类 DBS 植入手术、编程和报销类比 | 把 Parkinson's 症状治疗当作 BCI 需求 | 神经科服务线、Medicare、商业支付方 | 有用的报销和工作流类比,但临床任务不同 |
| 运动恢复邻近市场 | BCI 外骨骼、康复解码,以及脑到设备恢复项目 | 与神经接口无关的常规康复支出 | 研究型医院、康复中心、拨款 | 初始计算机控制用例之外的扩张向量 |
| 排除的广义 BCI 支出 | 游戏、军事、消费级头显、泛神经科技软件 | 没有可植入瘫痪恢复相关性的任何支出 | 消费者、国防、一般 R&D 买方 | 与叙事背景相关,但不计入 Neuralink SAM |
各行是边界逻辑,不是可相加的 TAM 桶。本表把 Neuralink 可植入辅助核心,与替代品、类比对象和邻近扩张类别区分开。
[CM001, CM002, CM005, CM010, CM011, CM012]从广义 BCI 收入,到受证据约束的可植入需求,再到 Neuralink 近期可触达市场,三层镜头拆开看。
这张图是概念框架,不是加总模型。它说明为什么广义 BCI 头条数字会高估 Neuralink 当前可服务市场。
[CM006, CM007, CM008, CM014, CM015, CM016]2.2 多视角测算、相互矛盾的估计与证据约束下的 SAM
公开测算数据确认 BCI 是真实市场,但不足以支撑一个可承销的 Neuralink TAM。本次审阅的分析师页面把 2026 年市场规模放在约 $295.5 million 到 $3.33 billion 之间;仅植入式来源和 Grand View 的理论侵入式 TAM 框架又把区间拉得更宽。这些估计互相冲突,因为它们混合了广义 BCI 收入、更窄的神经技术子集、仅植入式收入池和理论可触达机会。更安全的读法应看结构而不是标题数字:非侵入式系统仍占据多数已发布收入,侵入式系统今天更小但常被预测增长更快,北美仍是领先商业化区域。第二个视角来自疾病患病率。WHO 估计全球约 15.4 million 人带着脊髓损伤生活,CDC 则预计美国 ALS 患病人数将从 2022 年约 33,000 例增至 2030 年超过 36,000 例。这些人群池显示未满足需求有意义,但不能直接换算成 Neuralink SAM,因为候选资格仍取决于四肢瘫痪严重程度、手术适应性、语言和支持要求、站点可及性以及支付方覆盖。因此,本章保留规模区间,并明确把 Neuralink 特定的 SAM/SOM 留为未解决问题。[CM007, CM008, CM014, CM015, CM016, CM017]
| 发布方 / 视角 | 年份 | 地区 | 数值 | CAGR | 方法 / 单位 | 置信度 | 局限 |
|---|---|---|---|---|---|---|---|
| Fortune Business Insights | 2026/2034 | 全球 | 2026 年 $295.5M 至 2034 年 $960.8M | 15.9% | 广义 BCI 收入估计 | 中 | 远低于同业报告,边界可能更窄 |
| Mordor Intelligence | 2026/2031 | 全球 | 2026 年 $1.33B 至 2031 年 $2.69B | 15.08% | 神经科技 BCI 市场 | 中 | 不限可植入,并混合邻近模式 |
| Coherent Market Insights 估算 | 2026/2033 | 全球 | 2026 年 $2.75B 至 2033 年 $7.14B | 14.6% | 广义 BCI 市场 | 中 | 范围宽于 Neuralink 可植入用例 |
| Precedence Research | 2026/2035 | 全球 | 2026 年 $3.33B 至 2035 年 $13.86B | 16.77% | 广义 BCI 市场 | 中 | 受非侵入式和非医疗类别影响很大 |
| The Business Research Company | 2026/2030 | 全球 | 2026 年 $1.53B 至 2030 年 $2.52B | 13.3% | 可植入 BCI 市场 | 中 | 分类仍包括非侵入式和消费级邻近市场 |
| Grand View Research 理论视角 | 2025/2033 | 全球 | 侵入式 TAM $168.27B;非侵入式 SOM $397.59M | 1.52% / 8.73% | 理论 TAM 与 SOM 框架 | 低 | 不可与年度收入估计比较,也不应取平均 |
| 疾病患病率视角 | 2021-2030 | 全球 / 美国 | 全球 15.4M SCI 患者;美国 ALS 从 2022 年约 33k 增至 2030 年 >36k | n/a | 潜在候选人群,不是收入 | 高 | 不能直接转化为植入资格或有支付方支持的需求 |
本表有意保留不兼容的视角,而不是压成一个头部数字。已审阅公开来源没有隔离出 Neuralink 专属 SAM 或 SOM,并同时纳入候选资格、站点容量、定价和报销。
[CM006, CM007, CM008, CM014, CM015, CM016]已发布市场规模端点的区间,显示 BCI 规模测算会随边界定义大幅变化。
所有数值均为十亿美元,但第三行有意混合非侵入式 SOM、可植入 2026 年收入估计和理论侵入式 TAM,以说明这些来源不应被平均成一个承销数字。
[CM014, CM015, CM016, CM017, CM018, CM019]2.3 买方、用户、支付方与采用路径
买方地图比疾病人群池窄得多。今天的第一类买方是能赞助或承接植入研究、组织神经外科和康复流程并管理随访的学术医院或专科医院。日常用户是患者,但运营用户还包括负责校准和维护性能的神经外科医生、神经科医生、治疗师和 BCI 工程师。当前支付方通常是研究资金、公司资本、慈善资金或拨款支持的某种组合,而不是常规保险报销。如果品类成熟,之后的支付方很可能是 Medicare、商业保险公司,以及接受“手术 + 支持”经济性的医疗系统。这一点重要,因为 Neuralink 卖的不是一次性硬件盒子;它试图把托管式神经接口嵌入医院和家庭工作流。因此,采用路径从转诊和筛查开始,进入研究性植入,再到解码器训练和居家使用,最后才可能走向规模化报销路径。广义 BCI 报告中的公开医院份额数据支持以医院为中心的采用,但缺少植入特定报销意味着,在证据和编码更标准化之前,预算所有权会高度集中。[CM001, CM002, CM011, CM016, CM027, CM028]
| 细分 | 买方 | 用户 | 付款方 | 工作流 | 预算归属 | 采用触发因素 |
|---|---|---|---|---|---|---|
| 学术植入试验中心 | 主要研究者 / 神经外科服务线 | 患者、外科医生、BCI 工程师、治疗师 | 公司资本、资助金、慈善资金、研究预算 | 转诊、筛查、植入、解码器训练、居家随访 | 研究预算加医院能力分配 | 需要产出安全性和有用性证据 |
| 专科康复或神经恢复项目 | 康复科负责人 / 附属医院 | 患者、治疗师、康复人员 | 项目资金、资助金支持、未来付款方试点 | 结构化沟通或运动恢复方案 | 服务线运营预算 | 需要拿出超出标准护理、可衡量的功能改善 |
| 未来由付款方支持的瘫痪沟通服务 | 医院管理层及付款方合同支持 | 患者加多学科照护团队 | Medicare、商业保险、医疗系统 | 植入单次治疗加持续软件和支持管理 | 手术预算加长期照护预算 | 需要编码、覆盖范围和支付确定性 |
| 居家护理与辅助沟通替代路径 | 患者、照护者、言语团队、DME 供应商 | 患者和照护者 | Medicare DME、私人保险、自费 | 眼动追踪或 SGD 配置,配合远程沟通使用 | DME 预算或家庭预算 | 现在就需要风险更低的沟通辅助工具 |
| 相邻神经恢复平台 | 研究中心 / 创新资助赞助方 | 患者加外骨骼或机器人团队 | 资助金或赞助研究 | 概念验证型运动恢复工作流 | 研究和转化 R&D 预算 | 常规临床铺开前需要可行性数据 |
这里刻意把买方和付款方拆开。植入式 BCI 仍缺少常规报销,Neuralink 的早期市场会由医院和赞助方主导。
[CM002, CM011, CM027, CM028, CM029, CM030]矩阵展示 Neuralink 可能的早期采用细分市场及替代路径,在买方、付款方和工作流上的差异。
[CM009, CM010, CM011, CM020, CM027, CM028]从候选人识别到长期支持使用的价值链,标出商业化瓶颈出现的位置。
这条流程展示的是流程阶段,不是转化率。当前最大的瓶颈是筛查能力、报销和支持连续性。
[CM001, CM002, CM029, CM030, CM031, CM032]2.4 增长驱动、采用约束与保留的证据缺口
Neuralink 所在品类的增长逻辑很直接:瘫痪和沟通丧失存在严重未满足需求,已发表植入试验持续改善真实世界表现,相邻系统也显示运动和感觉的双向恢复正在推进。但门槛约束同样清楚。可植入 BCI 面临高风险医疗器械监管、手术和长期安全义务,商业化路径也比非侵入式替代方案难得多。GAO 和市场政策来源还强调试验结束后的结构性缺口,包括维护资金、数据所有权和用户支持。报销是近期最锋利的瓶颈。CMS 已经覆盖语音生成设备,医疗系统也长期拥有 DBS 的 Medicare 编码,但可植入 BCI 仍缺少专门的编码、覆盖和支付框架;Anthem 还明确把 BCI 康复设备归类为研究性且非医疗必要。因此,信任、植入后支持、资本强度和保险公司行为,对采用时点的影响至少不低于原始技术进展。合适的投资姿态是保留相互矛盾的规模估计,把需求按真实但分阶段来承销,并继续追问定价、站点容量、支持成本和支付方路径,而不是用一个宣传式 TAM 数字填补这些缺口。[CM003, CM004, CM013, CM021, CM022, CM023]
| 驱动因素 / 约束 | 方向 | 时间 | 含义 | 尽调问题 |
|---|---|---|---|---|
| 瘫痪和沟通领域存在大量未满足需求 | 需求驱动(+) | 当前且持久 | SCI 和 ALS 人群支撑了辅助控制的真实临床需求 | 量化有多少候选人同时满足手术和支持条件 |
| 已发布的用户表现优于传统 AAC | 需求驱动(+) | 当前 | 对部分用户来说,BCI 打字和居家使用可跑赢缓慢的眼动追踪基线 | 要求可比的真实世界 QoL 和弃用数据 |
| 北美医院和研究资源集中 | 需求驱动(+) | 近期 | 支持先在具备试验能力的中心商业化 | 梳理中心层面的站点容量和转诊网络 |
| 部分分析师模型显示侵入式细分市场增长较快 | 需求驱动(+) | 2026-2035 | 虽然基数小,但显示投资人兴趣和品类势能 | 检验增长假设是否隐含现实可达的植入吞吐量 |
| 高风险植入物的监管路径 | 约束(-) | 当前 | IDE、研究设计和上市后义务拖慢部署 | 审查临床终点、随访时长和制造控制 |
| CMS 没有专门的植入式 BCI 覆盖路径 | 约束(-) | 当前 | 缺少编码 / 支付清晰度,医院常规采购和患者可及性仍被卡住 | 尽早接触 CMS 路径和报销顾问 |
| 私人付款方持怀疑态度 | 约束(-) | 当前 | 即便 FDA 取得进展,Anthem 的试验性立场仍显示证据门槛很高 | 跟踪付款方政策变化和证据阈值 |
| 试验后支持和维护风险 | 约束(-) | 当前到长期 | 植入物是需要管理的系统,研究结束后仍需要资金 | 明确取出、维护和软件支持义务 |
| 信任、隐私和数据所有权 | 约束(-) | 持续存在 | 治理改善前,医院和患者可能会谨慎推进 | 评估神经数据权利、同意机制和网络安全态势 |
| 资本强度和多学科工作流负担 | 约束(-) | 持续存在 | 植入成功依赖手术、康复、工程和居家支持 | 假设规模化前,先建模单患者成本栈和人员负荷 |
这张表把每个增长论点都对应到一个运营阻碍或尽调问题。Neuralink 的市场可以有吸引力,但采用仍会受限。
[CM003, CM004, CM009, CM013, CM024, CM027]03竞争格局
3.1 按解决方案类别看格局,而不只是按创业公司名单
Neuralink 的相关竞争集合比一串高调植入式创业公司更宽。直接同业是那些试图为重度瘫痪人群恢复数字控制或沟通能力的可植入 BCI 公司,Synchron 和 Precision Neuroscience 是当前最清楚的例子。既有玩家也重要:Blackrock Neurotech 和 BrainGate 联盟拥有最深的已发表人体植入记录,哪怕它们距离打包的大众市场产品还更远,更接近研究基础设施和学术定制栈。Paradromics 应放在潜在进入者桶里,因为它追求的是同一项语音和电脑控制任务,只是采用更新的高带宽架构,并且到 2026 年才开始临床研究。 同样重要的是,买方不选择 Neuralink 也能解决很多用户需求。Tobii Dynavox 已经通过评估和资助流程销售带眼动追踪的语音生成设备;Medtronic 的 DBS 业务则给医院一个熟悉的植入式脑部手术和报销基准,尽管 DBS 解决的是不同临床任务。这意味着 Neuralink 同时在信号质量、手术负担、医院工作流、报销可信度和替代方案充分性上竞争。因此,竞争问题并不只是“谁的通道数最高”;真正的问题是,哪种方案能以可接受的风险、支持和经济性,跑通从手术室到居家使用的真实路径。[CP001, CP002, CP003, CP004, CP005, CP006]
| 类别 | 代表性解决方案 | 买方任务 | 为什么会约束 Neuralink |
|---|---|---|---|
| 直接植入同行 | Synchron | 为重度瘫痪人群提供长期数字控制 | 用摩擦更低的血管内手术切同一批核心用户 |
| 直接植入同行 | Precision Neuroscience | 医院部署的皮层记录,以及未来意念控制计算 | 把高带宽主张和创伤更小、可逆的手术叙事打包 |
| 既有平台 | Blackrock Neurotech | 研究级皮层内控制、刺激和解码 | 拥有最长的可见人体平台历史,也掌握该领域大量研究工具 |
| 学术机构自建 | BrainGate consortium 联盟 | 顶尖中心内定制的电脑控制和沟通神经假体 | 证明没有封装成商业产品,也能跑出顶级性能 |
| 潜在进入者 | Paradromics | 语音恢复和高速电脑控制 | 如果 2026 年试验数据验证临床前吞吐量,可能赢下沟通优先的细分市场 |
| 替代品 / 既有类比 | Tobii Dynavox SGD;Medtronic DBS 类比 | 无需手术即可沟通;已有报销的植入式脑部工作流 | 即便功能集不同,也在便利性、编码和医院经济性上设定采用门槛 |
这些类别只在战略层面互斥;实际决策时,买方可能在同一个周期里同时比较一家植入物创业公司、一个研究栈和一个非植入替代品。
[CP001, CP002, CP003, CP004, CP005, CP006]在植入对手中,Synchron 位于部署最容易的区域;Precision 兼具中等部署便利性和较强监管就绪度;Blackrock/BrainGate 仍是高能力但低规模路径。Tobii 锚定最容易部署的替代象限,而 Neuralink 和 Paradromics 聚集在高能力 / 高操作负担的一端。
X 轴 = 部署便利性和渠道就绪度(1 = 仅精英中心的定制堆栈,5 = 今天已可广泛采购)。Y 轴 = 控制丰富度和未来上行空间(1 = 狭窄症状或沟通用途,5 = 高带宽通用数字控制前景)。分数是基于公开手术描述、监管状态、报销成熟度和产品范围综合得出的有证据支持的序位评估;不是实测市场份额。
[CP002, CP003, CP004, CP006, CP007, CP008]3.2 直接同业、既有玩家与潜在进入者
Synchron 是最近的直接商业对手,因为它瞄准同一终端用户——需要数字自主能力的瘫痪人群——但产品策略完全不同。它的 Stentrode 设计为通过颈静脉和血管用导管送达,而不是开颅手术;公司也明确在为商业化、AI 解码和关键试验准备融资。Precision Neuroscience 是对常见 Neuralink 叙事最有力的监管反例:它的 Layer 7-T 已获得 FDA 510(k) clearance,可用于短期皮层使用;公司把自己定位在可逆性、医院规模和低创伤手术通路上,而不是最大化曝光。 Blackrock Neurotech 和 BrainGate 展示了这个领域的证据基础真正从哪里来。Blackrock 的 NeuroPort 平台拥有最长的可见人体研究足迹,并支撑了已发表皮层内文献的很大一部分;BrainGate 则继续在顶尖中心通过可行性研究用户和高速打字结果证明性能。Paradromics 是最重要的下一波进入者,因为它针对语音恢复优化,采用全无线、高数据速率系统,并在 2026 年启动临床研究。合在一起,这些竞争对手说明 Neuralink 面对的不是一个标准化单一对手;它面对的是一组由手术路径、监管路径、证据路径和语音场景分别主导的替代方案,分别攻击同一待完成任务的不同部分。[CP009, CP010, CP011, CP012, CP013, CP014]
| 公司 / 路径 | 类别 | 规模 / 融资信号 | 目标用户 | 产品范围 / 策略 | 关键限制 |
|---|---|---|---|---|---|
| Neuralink | 直接同行 / 参照点 | 知名度高、融资充足的私营临床阶段项目;仍属试验性 | 重度瘫痪人群,以及更长期、更广泛的 BCI 用户 | 机器人植入的高通道皮层内系统,目标是垂直整合 | 尚未获得完整商业批准;开颅工作流和支持经济性仍未解决 |
| Synchron | 直接同行 | 2025 年 $200M Series D;累计 $345M;公开点名 10 名植入患者 | ALS、SCI、中风及其他需要数字控制的重度瘫痪病例 | 血管内 Stentrode 加 AI 解码和 Apple 设备集成;为手术规模化优化 | 通道数低于顶级皮层内竞争对手;仍未商业化 |
| Precision Neuroscience | 直接同行 | $102M Series C;累计 $155M;公开引用至少 27-37 名研究患者 | 瘫痪用户加神经外科临床 / 研究环境 | 柔性 1,024 电极皮层薄膜,可逆放置;当前用于短期临床,之后拓展到更广泛的计算场景 | 当前 FDA 许可是临时 30 天使用,不是长期居家部署 |
| Blackrock Neurotech | 既有平台 | 2024 年获 Tether $200M 多数股权投资;公司投资者公告称已帮助 >40 人 | 研究型医院、神经假体实验室和重度残障植入研究 | NeuroPort / Utah Array 平台,覆盖多类临床研究任务的记录和刺激 | 商业化叙事仍是平台商业化,而不是常规广泛护理交付 |
| BrainGate | 学术机构自建 | 多中心 NIH / 学术可行性项目,不是风投支持的产品公司 | 顶尖研究中心内的四肢瘫痪用户 | 定制皮层内技术栈,经由顶级医院证明快速打字和电脑控制 | 不是封装产品,也无法通过普通医院采购部署 |
| Paradromics | 潜在进入者 | 临床研究阶段进入者,定位沟通优先;未公开定价 | 需要语音和电脑控制的 ALS、中风、SCI 或四肢瘫痪人群 | 围绕高速语音恢复打造的全无线 421 微电极系统 | 人体研究结果仍待公布,设备仍属试验性 |
融资和患者数字段结合了公司官方说法与独立报道,因为公开披露并不总是一致;这张表强调方向性规模,而不是股权结构表层面的精确数字。
[CP009, CP010, CP011, CP012, CP013, CP014]Synchron 在微创部署上领先,Precision 在公开监管进展上领先,Blackrock/BrainGate 在已发表证据深度上领先,Paradromics 的语音优先野心最强,Tobii 仍是仅沟通用途中最成熟的替代品。
[CP011, CP013, CP015, CP016, CP021, CP024]3.3 定价不透明、分销权力与切换成本
本章没有一家领先可植入 BCI 厂商公布清晰的长期植入价格表,这本身就是重要竞争事实。Synchron、Precision、Blackrock/BrainGate 和 Paradromics 仍通过研究、医院伙伴和研究项目提供访问,而不是通过普通采购页面。相比之下,Tobii 已经提供带眼动追踪、捆绑软件和资助流程的打包 SGD,可以通过治疗师、支付方或自费购买完成。CMS 的 SGD 规则展示了成熟覆盖是什么样:明确的 HCPCS 编码、文件标准和明示的不覆盖边界。Medtronic 的 2026 DBS 指南在植入侧展示了同一件事:具体医生支付锚点和耐久的专科渠道。 这种不对称构成了当下这个市场真正的分销护城河。医院系统、神经外科医生、康复团队和支付方对准入的控制远大于消费品牌。一旦患者植入,切换成本也异常高,因为硬件、手术团队、解码软件、训练数据和随访支持都绑在一起。多归属主要存在于植入层之上:用户可以比更换核心神经硬件更容易地组合不同软件端点或输出设备。因此,Neuralink 一旦完成植入就有一定锁定效应,但它尚未像 DBS 既有玩家或 Medicare 编码替代品那样控制上游渠道权力。[CP032, CP033, CP034, CP035, CP036, CP037]
| 解决方案 | 公开定价或报销信号 | 包含能力 | 触达患者 / 买方路径 | 对 Neuralink 的含义 |
|---|---|---|---|---|
| Neuralink | 未披露 | 试验性植入物、手术、解码器软件、随访 | 临床试验和医院路径 | 仅靠品牌无法消除采购和支持摩擦 |
| Synchron | 未披露 | Stentrode 植入物加数字设备控制和 Apple 集成 | 已点名试验站点,以及未来关键性 / 商业化医院铺开 | 手术摩擦更低,可能缩短站点采用周期 |
| Precision Neuroscience | 未披露 | 已获临时许可、用于脑图谱绘制的皮层接口;更长期的 BCI 路线图 | 医院 / 神经外科合作关系 | 对医院买方来说,监管进展可能比公关声量更重要 |
| Blackrock / BrainGate 技术栈 | 未公开;研究项目经济性 | 皮层内电极、定制解码技术栈、研究人员 | 顶尖学术中心和试验预算 | 最强证据基础仍依赖定制化运营模式 |
| Tobii Dynavox TD I-Series | 按报价、经资助方撮合或自购;Medicare 已有 SGD 编码 | 眼动追踪 AAC,捆绑软件和手机 / 网页 / 居家控制功能 | 治疗师评估、资助方材料,或直接购买 | 沟通优先的用户无需手术也能获得实质价值 |
| Medtronic DBS | 报销指南公开了 2026 年医生支付锚点 | 植入式神经刺激治疗,配专科支持和慢病照护工作流 | 成熟的医院、神经科和报销渠道 | 脑部植入采购在其他领域已经成熟,BCI 创业公司必须匹配这种严谨度 |
本次评估中的 BCI 创业公司都没有公开植入物定价。无法获得准确标价时,表格记录可获得的合同或报销信号,而不是猜测。
[CP032, CP033, CP034, CP035, CP036, CP037]3.4 护城河耐久性、替代风险与领域反向证据
这里的反向证据并不主要是某个具名对手已经赢下市场,而是整个品类在结构上仍未完成。Blackrock 的领域回顾和 Nature 的市场概览都指出,植入式 BCI 已进入人体试验数十年,却还没有一个完全获批的医疗市场赢家。GAO 更进一步,识别出悬而未决的数据权利、试验后支持和 CMS 覆盖问题,这些问题高于任何单家公司。它们不是抽象政策担忧:它们直接影响医院、支付方和家庭是否会信任一家永久植入物供应商来承担长期支持义务。 这让 Neuralink 的护城河真实存在,但还不耐久。Synchron 可以从低摩擦手术侧切入市场;Precision 可以从监管和医院侧切入;Blackrock 和 BrainGate 可以持续证明深度证据比品牌更重要;Paradromics 可以用更高吞吐量主张攻击语音细分;Tobii 式替代品则可以让很多以沟通为先的用户完全避开手术。Neuralink 仍受益于资本、注意力和激进的垂直整合野心,但领域证据显示,只要买方更看重更低侵入性、已知报销或已验证支持,而不是最大通道数,商品化和替代风险就已经可见。[CP041, CP042, CP043, CP044, CP045, CP046]
| Neuralink 护城河主张 | 威胁来源 | 严重性 | 证据 | 缓释措施 / 尽调问题 |
|---|---|---|---|---|
| 最高通道数保证产品领先 | Synchron 和替代品供应商靠更低摩擦部署取胜 | 高 | Synchron 公开用通道数换导管递送;Tobii 已经无需手术解决沟通 | 要求按手术负担调整后的真实世界任务基准,而不只是原始带宽 |
| 监管领先会创造赢者通吃的大部分市场 | Precision 已获得有意义的短期皮层使用 FDA 许可 | 高 | Precision 在这组公司里拥有最清晰的公开 FDA 里程碑,尽管还不是长期批准 | 询问 2026-2027 年哪些里程碑真正影响医院委员会和付款方 |
| 最深的证据基础属于 Neuralink | Blackrock / BrainGate 拥有最长的可见人体记录 | 中高 | Blackrock 支持的综述和 BrainGate 论文覆盖数十年植入研究和居家使用 | 将已发表性能和参与者支持负担,与 BrainGate 式证据对标 |
| 资本和公关声量足以锁定下游采用 | GAO 指出覆盖范围、维护和数据权利瓶颈,高于任何单一供应商 | 高 | 即便公司资金充足,领域层面的政策阻碍仍可能拖住采用 | 承销规模前,要求付款方、取出、质保和试验后支持计划 |
| 语音恢复细分市场以后自然会归 Neuralink | Paradromics 专门围绕沟通吞吐量打造 | 中高 | Connexus 针对合成语音、文本和电脑控制优化,计划 2026 年启动临床研究 | 跟踪首批人体语音结果和延迟,并与 Neuralink、BrainGate 结果对比 |
| 一旦植入,客户永远不会切换 | 植入锁定效应真实存在,但上游分销权仍掌握在医院和付款方手中 | 中 | DBS 和 SGD 替代品表明,转诊、编码和支持系统决定谁一开始能获得植入 | 承销渠道合作和医院经济性,不只看植入后留存 |
严重性评级是定性判断,以证据为先。登记表关注的是:在任何单一供应商成为主导者之前,竞争性手术类别、监管姿态和替代品会从哪些位置侵蚀 Neuralink 的差异化。
[CP041, CP042, CP043, CP044, CP045, CP046]这个领域竞争激烈,因为不同对手各自掌握不同就绪优势:Synchron 在手术路径上占优,Precision 在公开 FDA 进展上占优,Blackrock/BrainGate 在证据深度上占优,Tobii 在可部署替代品经济性上占优,GAO 定义的政策缺口则构成下行压力。
[CP009, CP015, CP024, CP028, CP035, CP042]3.5 图表
04财务情况
4.1 收入模式、定价不透明与公开牵引力缺口
公开记录支持一个连贯的未来收入机制,但不支持商业收入判断。Neuralink 披露的产品路径是面向重度瘫痪人群的机器人植入 BCI,仍在通过研究性试验和医院伙伴推进,而不是进入普通采购。官方和独立的 2025 年 6 月融资报道称,已有五名植入用户活跃,四项 Telepathy 试验正在运行;Series E 公告称资金将用于扩大患者准入和未来设备。缺失的是投资级桥梁:这些里程碑如何转化为确认收入。保留下来的来源没有披露标价、实际报销、合同期限、支持绑定、收入组合或确认销售额。因此,收入质量是本章核心未知数。商业模式看起来更接近高触达、由医院和支付方中介的植入平台,而不是纯软件订阅;但本章必须把定价、收入确认和变现组合视为私有证据缺口,不能用编造数字填补。[CI001, CI002, CI003, CI004, CI005, CI006]
| 来源 | 机制 | 单位 | 当前价值 / 状态 | 收入质量 | 尽调问题 |
|---|---|---|---|---|---|
| Telepathy 植入套装 | 植入硬件加机器人辅助手术和激活 | 手术 / 植入单次治疗 | 未披露公开商业销售;证据仍是临床试验使用 | 商业报销和回款出现前仍未证实 | 要求首个站点合同、开票到回款瀑布,以及收入确认政策 |
| 植入后编程和解码器支持 | 编程、算法调优、随访和维护 | 就诊 / 服务单次治疗 | 从品类类比看经济上可行,但 Neuralink 没有公开价格或报销表 | 可能成为经常性收入,也可能劳动密集、稀释利润率 | 要求支持人员模型、报销假设和植入后服务毛利率 |
| 辅助设备 / 机器人扩展 | 屏幕控制之外,未来 CONVOY 式使用场景 | 附加项目 / 适应症 | 仅处于试验阶段 | 是可选性,不是当前收入 | 要求合作伙伴经济性、附加率假设和增量支持负担 |
| 视觉和语音恢复项目 | 在 FDA 突破性器械认定下,未来扩展适应症 | 平台 / 适应症扩展 | 商业化前 | TAM 扩大可能先推高 R&D 烧钱,再改善变现 | 按适应症要求预算、时间线和预期报销路径 |
| 研究或赞助研究经济性 | 潜在资助金、站点报销或研究支持 | 项目资金 | 留存来源没有披露赞助研究收入来源 | 无法基于公开数据承销 | 要求试验预算归属、赞助方报销和站点付款条款 |
| 收入结构披露 | 硬件、支持、软件、服务之间的拆分 | 收入结构 | 未披露 | 判断耐久性或经常性内容的阻塞问题 | 要求按来源拆分的月度收入桥,以及按来源确认的毛利率 |
各行把可能的变现机制和已披露商业证据拆开。Neuralink 没有发布定价或已确认销售数据时,表格明确记录缺口,而不是猜测。
[CI001, CI002, CI003, CI004, CI017, CI036]| 要素 | 价格 / 合同信号 | 标价与实际价格 | 当前公开解读 | 来源基础 |
|---|---|---|---|---|
| 植入硬件 + 手术套装 | 未披露 | 没有公开标价或实际价格 | 留存的 Neuralink 和新闻来源讨论融资和试验,不讨论患者定价 | Neuralink 官方 + CNBC 报道 |
| 临床试验准入 | 入组和筛选,不是开放商业购买 | 不是商业价格信号 | 现阶段准入由研究标准和站点参与共同控制 | Barrow PRIME + ClinicalTrials 登记库 |
| 编程 / 随访 | 成熟脑设备类比已有独立医生支付代码 | 仅作类比;不是 Neuralink 已实现定价 | DBS 编码显示,植入后随访在经济上是独立服务 | Medtronic DBS 指南 |
| 医院设施经济性 | 成熟类比已有门诊和住院编码 / 支付路径 | 仅作类比 | 植入式神经业务大规模铺开前,经济性先取决于设施编码脚手架 | Medtronic DBS 指南 + CMS / MedPAC 背景 |
| 软件更新 / 算法再训练 / 康复 | Neuralink 未公开费率或代码 | Unknown | BCI 支持很可能包含持续服务,普通设备标价会漏掉这部分 | Inside BCI + GAO |
| 未来视觉 / 语音 / CONVOY 项目 | 未公开定价或支付方策略 | Unknown | 适应症扩张已经公开;商业化设计没有公开 | Neuralink + PharmaPhorum |
本表有意把 Neuralink 自身未知项和成熟植入式神经类比放在一起,让读者看清商业不透明卡在哪里:问题不是支持服务和设施经济性是否重要,而是 Neuralink 准备如何定价、如何回款。
[CI003, CI011, CI016, CI017, CI018, CI019]公开证据支持一条由医院居中串起的路径:从患者筛查走向未来可报销的植入收入,但价格和回款环节仍未披露。
这座桥是结构性的,不是数字模型,因为留存公开资料包没有披露 ASP、付款方组合或回款。
[CI001, CI002, CI012, CI016, CI017, CI018]4.2 GTM 动作与销售效率代理指标
Neuralink 当前市场进入动作最适合理解为医院站点创建、患者筛查和未来支付方赋能,而不是客户经理吞吐量。Barrow 的 PRIME 页面招募四肢瘫痪或 ALS 成人进入机器人辅助植入研究;官方 Series E 公告点名三个国家、两大洲的神经外科机构。FDA 指南解释了为什么这个动作缓慢且程序化:高风险神经植入物要经过预提交工作、IDE 支持的研究和之后的上市申请;Medtronic DBS 等成熟脑设备类比产品已经依赖具体的 CPT、HCPCS、医院和编程编码。这些事实足以推断销售和导入周期很长、临床教育负担重、部署需要大量支持。它们不足以计算 CAC、回本周期、转化率或站点级贡献毛利。因此,公开 GTM 证据是方向性的:Neuralink 看起来像一个 B2B2C 医疗器械推广,成败取决于医院委员会批准、外科医生工作流和报销执行,而不是仅靠消费者需求。[CI012, CI013, CI014, CI015, CI016, CI017]
| 指标 / 代理项 | 数值 | 置信度 | 重要性 | 尽调问题 |
|---|---|---|---|---|
| 已披露植入用户 | 5 | 高 | 能证明技术牵引力和早期用户验证,但不能证明付费需求或单位经济性 | 要求按站点拆分患者漏斗:从转诊到植入,再到活跃使用 |
| 已披露试验版图 | 四项 Telepathy 试验;官方全球站点横跨三国、两大洲 | 中 | 说明 GTM 靠站点扩张和临床运营,不是广泛商业吞吐 | 要求站点入驻周期、激活成本和站点级植入节奏 |
| 公开付费客户数 | 低 | 没有签约站点或已开票患者,就无法直接计算 CAC 或回本周期 | 要求披露签约医院数、已开票植入量和已回款理赔数 | |
| 已实现 ASP / 报销 | 低 | 没有已实现价格,就搭不起可信的收入或毛利桥 | 要求按植入队列披露开票金额、支付方结构和净回款率 | |
| 编程 / 支持人力负载 | 类比中的支持负担可见;Neuralink 自身工时不可见 | 中 | 服务很重的照护路径会吃掉看似漂亮的硬件或软件利润率 | 要求每位患者编程工时、再训练频率和现场临床团队人数 |
| 渠道组合 | 由医院站点、外科医生和支付方共同中介 | 中 | 意味着长周期、委员会式转化,而不是自助式增长 | 要求转诊来源、医院委员会时间线,以及任何分销商 / 合作伙伴经济性 |
本表使用公开代理项,因为 Neuralink 不披露获客成本、回本周期、留存或已实现定价。null 是有意保留的尽调卡点,不是零值。
[CI009, CI010, CI011, CI015, CI016, CI017]公开的单位经济性桥从新融资开始,止于仍未解开的毛利输出:最重的成本节点看得见,但实际价格看不见。
这张图有意停在未解输出,因为缺少价格、回款和服务成本数据,任何公开毛利估计都会是编造。
[CI005, CI025, CI032, CI033, CI034, CI035]4.3 成本结构、毛利路径与服务交付负担
可能的成本栈比头条患者数量里程碑暗示的重得多。Neuralink 的模式组合了植入硬件、专门制造、机器人辅助神经外科、站点导入、监管运营和反复发生的植入后支持。GAO 和 Inside BCI 把这一品类的服务负担说得异常清楚:可植入 BCI 可能需要长期维护、软件更新、算法再训练,以及试验阶段后的支持;一些历史参与者已经在资金消失后失去访问。公开类比也指向同一方向。Medtronic 的 DBS 报销指南显示,植入式脑部流程包含独立的植入、发生器和编程活动;Medtronic 的 2025 10-K 则显示,成熟植入物业务线会背负多少制造、上市和现场商业化支出。学术 BCI 商业化回顾更让人警醒:即便一个植入项目在科学上成功,如果制造成本、康复负担、报销摩擦和医生采用不能同步放大,商业上仍可能失败。如果 Neuralink 最终把固定成本摊到更多植入上,长期毛利率可能改善;但近期毛利路径资本密集、支持负担重。[CI018, CI019, CI020, CI022, CI023, CI024]
| 成本模块 | 公开证据 | 毛利影响 | 营运资本 / 资本开支影响 | 尽调问题 |
|---|---|---|---|---|
| 植入硬件和专用组件 | BCI 商业化文献强调,新型植入物制造成本高,且依赖专用组件 | 规模和良率改善前,COGS 压力高 | 库存、报废和供应商集中风险 | 要求 BOM、良率、供应商条款和质保假设 |
| 机器人辅助植入流程 | Barrow 描述了机器人辅助手术;成熟类比仍需要多个带编码的流程步骤 | 近期手术成本高 | 资本设备、培训和站点就绪负担 | 要求机器人利用率、折旧政策和单站点搭建成本 |
| 临床试验网络和监管运营 | 官方材料强调全球试验;FDA 路径需要 IDE,之后还要提交上市申请 | 收入前运营开支 | 大规模销售前持续烧钱 | 要求单站点试验预算、监查成本和监管团队人数 |
| 植入后编程、软件更新和再训练 | Inside BCI 和 GAO 指出持续软件和支持需求 | 即使硬件利润率改善,服务人力也会压缩毛利 | 长周期支持可能需要计提 / 储备 | 要求支持储备政策、随访节奏和报销地图 |
| 试验后维护 / 取出义务 | GAO 记录过试验结束后支持资金断裂的历史案例 | 带来尾部负债和声誉风险 | 可能需要资产负债表储备或赞助方承诺 | 要求试验后支持资金计划和取出责任矩阵 |
| 商业发布和医院签约 | Medtronic 10-K 显示,成熟植入物业务存在大额 SG&A 和发布驱动的商业化开支 | 批准后,现场商业化开支仍可能保持高位 | 销售团队、教育和报销支持会吸收营运资本 | 要求发布预算、现场临床组织设计和报销支持支出 |
各行把 Neuralink 自身公开信号和明确标注的植入式神经类比放在一起。目的不是倒推出一个虚假的毛利率,而是指出利润率压力最可能从哪里冒出来。
[CI018, CI019, CI023, CI024, CI025, CI026]Neuralink 的现金需求集中在新型制造、站点搭建和长期支持义务上;而公开可见度最弱的位置,正是承销最需要数据的位置。
该矩阵为定性判断,因为公开来源只描述成本类别和政策瓶颈,并未披露 Neuralink 的实际预算、储备或烧钱数值。
[CI018, CI020, CI023, CI024, CI025, CI033]4.4 资本充足性、融资依赖与财务结论
融资图景明显强于收入图景。官方披露支持 2025 年 6 月 $650 million Series E 和 2023 年 8 月 $280 million Series D,这给出过去两轮具名融资中至少 $930 million 的清晰公开底线。它降低了短期破产风险,也显示公司能反复获得资本,但没有回答承销真正关心的问题:保留下来的公开来源没有披露账上现金、月烧钱速度、债务、营运资金需求或现金跑道月数。因此,下一次融资触发点是隐含的,而不是明示的。如果报销路径、站点经济性或从临床到商业的转化比预期更慢,Neuralink 很可能在公开投资者看到规范化植入毛利之前再次回到市场。财务上,结论是混合偏谨慎:公司披露的外部背书足以继续推进试验和产品扩张,但收入质量未验证,毛利形状不透明,资本强度高。最大的尽调阻碍是实际定价、报销设计、试验后支持准备金,以及完整的现金 / 烧钱 / 债务桥。[CI005, CI006, CI007, CI008, CI034, CI035]
| 项目 | 公开证据 | 当前数值 / 状态 | 含义 | 尽调问题 |
|---|---|---|---|---|
| 最近披露融资 | 官方 Series E 加 CNBC 佐证 | $650M Series E 于 2025-06-02 完成 | 显著延长运营灵活性,也验证了投资人渠道 | 确认交割时收到的非受限现金,以及是否有专项用途 / 分期承诺 |
| 近期公开披露股权融资下限 | 官方 Series D + Series E | 2023-2025 年具名轮次合计 $930M | 给出干净的公开资本下限,但不是在手现金 | 将公开轮次总额与股权结构表、现金桥和优先股堆叠对齐 |
| 在手现金 | 保留公开材料未披露 | 无法用公开证据计算现金跑道 | 要求最新资产负债表和月度现金滚动表 | |
| 月度烧钱速度 | 保留公开材料未披露 | 下一轮时点和下行韧性仍不透明 | 要求过去 12 个月总烧钱和净烧钱,并拆分为 R&D、临床、制造和 G&A | |
| 债务 / 项目融资义务 | 未保留到公开披露 | 外部承销人看不到潜在契约或还款请求权 | 要求债务明细、契约、抵押品,以及任何设备租赁或担保 | |
| 下一轮触发点 | 由报销和商业化时点隐含,而非公开说明 | 取决于里程碑 | 如果编码、覆盖、站点吞吐或利润率爬坡落后,规范化经济性可见前,公司很可能还要再融一轮 | 要求管理层情景计划,按植入量、支付方覆盖和现金阈值列出触发点 |
本表把近期融资事实与现金、现金跑道未知项分开。公开募集资本不等于多年 R&D 和临床扩张之后仍可动用的流动性。
[CI005, CI006, CI007, CI008, CI035, CI040]| 缺失指标 | 公开材料状态 | 重要性 | 具体尽调路径 |
|---|---|---|---|
| 已确认收入 / ARR / 商业销售 | 未披露 | 没有已确认销售,就没有收入质量、增长或估值锚 | 要求按适应症、地域和客户类型拆分月度已确认收入桥 |
| 按硬件 / 支持 / 软件拆分的毛利率 | 未披露 | 委员会无法拆分平台上行与手术、服务拖累 | 要求按植入硬件、手术支持、编程和软件服务列出毛利率瀑布 |
| 已实现 ASP 和报销瀑布 | 未披露 | 价格不透明会堵住所有单位经济性工作 | 要求按队列披露开票金额、支付方报销、折让、拒付和现金回款 |
| 营运资本、库存和支持储备 | 未披露 | 即使技术可行,植入物业务也可能败在库存、质保或支持义务上 | 要求库存周转、储备政策、取出义务和支持负债会计处理 |
| 债务、契约、租赁或表外义务 | 未披露 | 资本充足性和下行风险不能盲承销 | 要求债务明细、租赁义务、担保权益和契约余量 |
| 站点级单位经济性和报销设计 | 只有类比公开 | 医院是否采用,取决于站点和支付方能否让手术在经济上跑通 | 要求站点 P&L、人员配置模型、编码计划和植入后支持报销策略 |
这是审阅 26 个保留来源后价值最高的缺口。问题按投资人可直接使用的口径写成,方便管理层直接补上最大的承销洞。
[CI003, CI004, CI018, CI019, CI035, CI036]4.5 图表
05产品与技术
5.1 产品定义与公开模块地图
Neuralink 的产品不是一颗独立芯片,也不是通用 AI 接口。按工作流看,当前提供的是一项研究性脑机接口服务:先筛查患者,再做神经外科植入,随后训练解码器,并让重度运动障碍人群持续使用电脑控制。官方 Device Control 材料把近期用户定义为因脊髓损伤或 ALS 而双手使用受限或完全无法使用的人;Telepathy 更新则把待完成任务定义为恢复跨电脑、手机,并最终延伸到机械肢体的数字自主能力。因此,公司更接近一套托管式神经技术栈,而不是一次消费电子发布。 公开模块地图现在有三条可见产品线。第一是 Telepathy/PRIME,即用于光标和日常数字使用的核心设备控制路径。第二是 CONVOY,这是一个可行性扩展,用同一枚 N1 植入物控制研究性辅助机械臂,因此把产品从数字自由扩展到物理任务执行。第三是 Blindsight,这是未来视觉假体项目,绕开自然的眼到脑路径,把路线图推向感觉恢复。围绕这些模块的支持资产包括患者登记、医院站点网络、把信号解码为动作的 Neuralink Application,以及通过招聘和 GitHub 信号可见但有限的开发者 / 运营足迹。因此,产品家族最好理解为拥有多条临床适应症轨道的整合平台,而不是单个一次性植入物。[CE001, CE002, CE003, CE004, CE005, CE015]
| 模块 / 资产 | 主要用户 | 当前状态 | 差异化 | 主要尽调缺口 |
|---|---|---|---|---|
| Telepathy / PRIME 设备控制 | 使用电脑和手机的瘫痪成人 | 活跃的多站点研究项目 | 植入物 + 解码器 + 站点流程一体化,服务数字自主 | 未公开商业定价、耐久性曲线或支持经济性 |
| CONVOY 机械臂控制 | 需要物理任务辅助的现有或未来 PRIME 参与者 | 2024 年启动可行性扩展 | 复用 N1 植入物控制机械臂,而不是只控制光标 | 未公开独立性能数据或长期可用性结果 |
| Blindsight 视觉假体 | 视觉皮层完好、存在重度视力损害或失明的人群 | 未来美国试验处于登记阶段,已获 Breakthrough Device Designation | 绕过眼睛 / 视神经通路,直接写入负责视觉的脑区 | 本轮未公开人体有效性或长期安全性数据 |
| Patient Registry | 潜在参与者和照护者 | 活跃 intake 机制 | 让 Neuralink 在未来试验前收集结构化需求和适应症反馈 | 未公开转化、优先级排序或流失数据 |
| Neuralink Application / 解码层 | 植入接受者和临床团队 | 已上线,但披露不足 | 把无线神经数据转成可执行的电脑或设备命令 | 未公开架构、API 或网络安全文档 |
| 医院 / 试验站点网络 | 神经外科和康复合作伙伴 | 正在美国和英国站点扩张 | 让产品成为一种运营模型,而不只是硬件 | 未公开培训、维护或修正照护的服务水平指标 |
状态标签反映公开项目成熟度,而非监管批准。尽调缺口单元格有意保留公开证据中仍未披露的内容。
[CE001, CE003, CE004, CE005, CE020, CE022]| 用户任务 | 当前工作流 | Neuralink 方案 | 公开可见收益 | 关键限制 |
|---|---|---|---|---|
| 基础电脑和智能手机控制 | 手术、校准、监督式上手,然后独立日常使用 | Telepathy / PRIME | 仅凭思维恢复光标和设备控制 | 需要植入手术和专科随访 |
| 高频数字工作或创作任务 | 使用解码器和应用持续训练 | 控制能力扩展后的 Telepathy | 第二位参与者上手后很快使用 CAD 软件和游戏 | 详细生产力基准仍由公司报告 |
| 辅助性物理交互 | 从设备控制研究交叉入组 | CONVOY 机械臂可行性研究 | 把产品从数字控制延伸到物理自由 | 仍处可行性阶段,未发布长期结果 |
| 重度失明的视觉恢复 | 未来试验启动前先做登记入组 | Blindsight 视觉假体 | 通过刺激视觉脑区,瞄准没有可用眼睛 / 视神经通路的用户 | 本轮未公开人体表现数据 |
| 全球参与者获取和适应症扩张 | 登记库触达 + 医院站点选择 | Patient Registry + 站点网络 | 搭建未来试验漏斗并收集需求 | 未公开入组漏斗质量或地域公平性数据 |
收益仅限公开声明或站点报告的能力。本表不假设已广泛商业可得。
[CE002, CE003, CE005, CE018, CE020, CE021]当前产品工作流从登记和筛查开始,经过植入、解码器训练,最终进入日常外部设备使用。
[CE002, CE003, CE018, CE020, CE021, CE022]Neuralink 的公开能力图里,设备控制基础设施最强;已发表的信任、支持和未来适应证证据较弱。
定性成熟度评级概括截至本次运行日期的公开记录;并非公司发布的评分。
[CE001, CE004, CE020, CE022, CE023, CE032]5.2 架构、植入栈与运营模式
对仍处于研究阶段的 BCI 公司来说,这套架构的垂直整合程度异常高。Neuralink 技术页描述的是一款完全可植入、外观不可见的设备,包含密封生物相容外壳、无线感应充电、自研低功耗电子器件、高柔性线体,以及把神经信号解码为意图动作的 Neuralink Application。运行闭环是:从与运动相关的脑区记录神经活动,设备端电子器件处理并无线传输信号,解码器把信号转成指令,用户控制光标、手机或其他外部设备。公开页面始终把植入物、软件和用户体验呈现为单一产品系统,而不是可分开的 SKU。 第二根支柱是手术机器人,以及背后的制造和封装诀窍。Neuralink 自己的技术页称,线体细到人手无法操作;2019 年 JMIR 论文及其 PMC 版本则用具体工程语言描述了早期研究平台:96 根聚合物线体上最多 3,072 个电极,机器人每分钟插入六根线体,自研 ASIC 做低功耗放大和数字化,封装小于 23×18.5×2 mm3。同一篇论文也说明,研究平台用有线方式服务原始带宽开发,而未来临床设备预计转向密封全植入封装,配备无线供电和遥测。涵盖计算机视觉机器人、片上网络神经数据和聚合物植入物外壳的专利标题进一步说明,差异化不只在一颗芯片,而在线体、机器人、信号处理和封装栈的组合。[CE006, CE007, CE008, CE009, CE010, CE011]
| 层级 / 组件 | 作用 | 关键依赖 | 差异化信号 | 主要风险 |
|---|---|---|---|---|
| 柔性线阵列 | 记录目标脑区的神经活动 | 线体完整性和放置精度 | 超薄柔性线旨在尽量减少损伤 | 长期稳定性和线体迁移仍是核心风险 |
| 密封植入物外壳 | 在生理环境中保护植入物 | 封装材料和密封可靠性 | 生物相容性密封外壳是全植入设计的核心 | 未公开现场故障或更换率 |
| 电池和供电路径 | 通过无线感应充电为植入物供电 | 充电器人体工学和热管理 | 无线日常使用是用户独立叙事的核心 | 电池寿命和真实世界维护披露仍不足 |
| 定制芯片和设备端电子器件 | 放大、数字化并传输神经信号 | ASIC 设计、功耗预算和固件质量 | 低功耗定制硅和信号处理密度是核心护城河要素 | 公开固件、冗余和故障安全细节有限 |
| 手术机器人 | 准确高效地放置线体 | 光学、OCT、运动控制、耗材和外科医生工作流 | 机器人能完成徒手无法达到的放置,并支持避开血管 | 流程复杂度仍造成临床规模化瓶颈 |
| 解码软件 / Neuralink Application | 把信号转换为光标或设备命令 | 训练数据、算法调优和 UI 迭代 | 直接掌控解码与用户体验闭环,加快迭代 | 网络安全、日志和第三方集成细节没有公开 |
架构分析把当前公开产品页和早期技术论文放在一起,用来区分稳定的技术栈要素和仍然私有的实现细节。
[CE006, CE007, CE008, CE009, CE010, CE011]Neuralink 公开栈把植入硬件、机器人植入、解码软件和试验运营拼在一起。
该栈由官方产品、更新和技术文档来源综合而成;内部子系统的精确边界并未完全公开。
[CE006, CE008, CE010, CE011, CE012, CE020]Neuralink 的产品执行需要临床站点、手术机器人、封装、数据治理和试验后支持持续对齐。
依赖关系反映公开运营模型,而非已披露的内部组织图。
[CE018, CE025, CE030, CE031, CE034, CE035]5.3 部署、集成、支持、可靠性与路线图
今天的部署模式明确由医院主导,且支持负担很重。PRIME 招募通过 Barrow 和 University of Miami 等站点伙伴推进;UCLH 的 GB-PRIME 更新显示英国也是同样模式:患者接受筛查,在专门神经外科中心植入,学习使用数字设备,随后通过持续训练时段加独立日常使用获得支持。Miami 第五名参与者公告和 Neuralink 自有更新合在一起,展示了一个实际术后流程:一天内出院、初始导入、以及反复软件或任务调优。这不是即插即用可穿戴设备;它依赖手术、站点能力、照护者可靠性和持续随访。 可靠性进展真实存在,但仍很早期。Neuralink 第二名参与者更新公开称,第一名参与者经历线体回缩,性能一度下降;同一篇文章称,公司通过减少脑部运动、缩小植入物与脑表面的间隙来改变手术技术,并且在第二名参与者身上没有观察到线体回缩。这是令人鼓舞的迭代证据,但仍是公司报告且试验规模很小。路线图扩张也快于已发表证据基础。根据 Neuralink 2026 年 1 月更新,Telepathy 已达到 21 名参与者;CONVOY 把同一植入物延伸到机械臂控制;Blindsight 则在 FDA Breakthrough Device Designation 下增加视觉恢复路径。这些动作提高了期权价值,也传递平台野心,但在任何广泛商业发布可见之前,它们也扩大了支持、临床运营和长期可靠性的负担。[CE015, CE016, CE017, CE018, CE019, CE020]
| 日期 / 阶段 | 功能或里程碑 | 状态 | 含义 | 来源依据 |
|---|---|---|---|---|
| 2019 研究平台 | 3,072 通道柔性线平台与机器人植入论文 | 已发表的临床前基础 | 奠定了后续产品主张依托的密度、机器人和 ASIC 架构 | JMIR / PMC 论文 |
| 2024 PRIME / Telepathy 真实使用 | 首位参与者及后续用户体验里程碑 | 活跃且在扩展 | 验证了电脑控制工作流和日常使用叙事 | Neuralink 更新 + 站点合作方文章 |
| 2024 CONVOY 启动 | 借助 N1 植入物控制辅助机械臂 | 可行性研究 | 把路线图从数字自主扩到实体任务辅助 | CONVOY 官方启动 + Device Control 页面 |
| 2024 Blindsight 认定 | FDA 授予视觉假体突破性器械认定 | 试验前加速信号 | 开出第二条适应证路径,获得监管关注,但还不是批准 | 官方 Blindsight 更新 + MedicalDeviceNetwork |
| 2025-2026 站点扩张 | Miami 第五位参与者、UCLH 七人 GB-PRIME 队列、2026 年 1 月前宣布 21 名参与者 | 持续铺开 | 显示运营扩张和更快迭代,但也带来更高支持与质量负担 | Miami / UCLH / Telepathy 更新 |
路线图表把技术基础、已激活功能和未来适应证扩张拆开,方便读者分清哪些已经发表,哪些仍是愿景。
[CE013, CE015, CE016, CE017, CE018, CE019]5.4 差异化、信任、隐私、合规与未解决控制
Neuralink 的差异化在硬件、软件、手术和临床迭代交汇处最强。公开技术论文、专利线索和官方产品页面合在一起,支持公司在高密度柔性线体、自研电子器件、机器人辅助放置,以及直接拥有解码器和用户体验栈上拥有真实护城河。公司似乎也在搭建神经技术商业化所需的外围操作系统:由倡导组织参与设计的患者登记、覆盖工程 / 科学 / 运营的跨学科招聘,以及通过 GitHub 组织释放出的至少温和公开开发者信号。与强调单一组件的同业相比,无论是支架路径、皮层薄膜还是研究工具,Neuralink 的公开故事都异常端到端。 同样的端到端野心也制造了信任负担。Neuralink 隐私政策称,临床试验处理受协议、知情同意和适用时 HIPAA 授权约束;公司收集敏感健康和沟通数据;可能与研究伙伴和服务提供者共享数据;不出售个人信息,也不为定向广告共享个人信息。但独立来源仍提出实质担忧。GAO 警告,BCI 用户面临数据所有权不清、试验后可能失去访问或支持的问题。KU Leuven 和 Frontiers 认为,侵入式 BCI 带来异常严肃的自主性和神经隐私风险;Frontiers 还提示长期安全不确定性,以及围绕试验注册规范的既有透明度批评。PCRM 的记录和倡议页面让动物福利争议继续成为负面声誉包袱。因此,本章结论是:Neuralink 的产品边缘可信,但信任栈在公开证据中只得到部分证明;质量体系细节、网络安全设计、取出结果和长期支持义务仍严重披露不足。[CE024, CE025, CE026, CE027, CE028, CE033]
| 控制项或担忧 | 公开状态 | 范围 | 公开证据说明 | 剩余缺口 |
|---|---|---|---|---|
| 临床同意 + 方案治理 | 已披露 | 试验和登记库 | 隐私政策称,临床试验处理在适用情况下受方案、知情同意和 HIPAA 授权约束 | 方案细节和审计发现未公开 |
| 个人数据出售 / 广告共享 | 反向披露 | 网站和更广泛个人信息处理 | 隐私政策称 Neuralink 不出售个人信息,也不为定向广告共享个人信息 | 未公开用于执行或访问日志的技术架构 |
| 研究合作伙伴数据共享 | 已披露 | 医疗服务方、机构、学者、出版商、供应商 | 隐私政策允许与研究合作伙伴、医疗服务方和服务提供商共享 | 未公开合作伙伴名单或数据最小化指标 |
| 试验后准入和支持 | 独立担忧 | 植入式 BCI 用户整体 | GAO 警示,部分参与者在试验后失去准入或支持,数据所有权也不清晰 | 未公开 Neuralink 自身支持储备或取出政策 |
| 神经隐私和自主性 | 独立担忧 | 侵入式 BCI 治理 | KU Leuven 和 Frontiers 称,BCI 带来异常敏感的隐私和自主性风险 | 除通用隐私政策外,未公开神经权利或神经数据专门政策框架 |
| 动物福利和透明度包袱 | 反向 | 临床前信任和声誉 | PCRM 持续公开详细记录和指控,Frontiers 也提到透明度批评 | 目前没有公开第三方材料,把临床前争议与当前信任叙事做过对账 |
这张表把已披露控制措施和独立风险信号放在一起,因为质量体系的公开证据仍然稀疏。「公开状态」只说明外界看得见什么,不代表控制措施已经足够。
[CE033, CE034, CE035, CE036, CE037, CE038]5.5 图表
06客户情况
6.1 当前客户界面与分层
Neuralink 当前的“客户”最好理解为一个临床采用网络,而不是付费企业账户或已报销的医疗系统买方。当前终端用户是重度运动障碍者,他们想更直接地控制电脑、手机,并最终控制机械肢体。直接运营买方不是刷卡消费者,而是研究运营方:Neuralink 自身、伙伴医院、神经外科医生、试验协调员、照护者和监管者,他们决定患者能否被筛查、植入、支持和随访。Barrow 的 PRIME 页面、Miami 的站点公告和英国 GB-PRIME 材料都描述了一条严格筛选路径,包含年龄、伤情稳定性、沟通和照护者要求,使采用更像复杂临床导入,而不是广泛消费发布。 因此,今天的分层沿着几条轴展开。按用户病况,正在运行的 PRIME 通道聚焦颈髓损伤和 ALS;按地域,公开站点证据覆盖 Phoenix、Miami,以及由 UCLH 和 Newcastle 牵头的英国网络;按用例,基础产品是数字设备控制,CONVOY 则把同一植入用户延伸到辅助机械臂控制。支付方实际上仍是赞助方和试验栈,因为公开来源没有显示常规保险报销或医院采购项目。Neuralink 的登记系统把未来需求映射扩展到截瘫、视力丧失、耳聋和失语,但该登记是受理和研究渠道,不是已部署商业账户的证据。尽调上,关键区别是用户需求看起来宽于当前治疗患者数量,而渠道准备度仍取决于少数专门站点和协议驱动的基础设施。[CU001, CU002, CU003, CU004, CU005, CU026]
| 细分群体 | 买方 / 运营方 | 用户 / 受益人 | 付款方 / 资金来源 | 地域 / 渠道 | 当前证据和主要缺口 |
|---|---|---|---|---|---|
| PRIME 瘫痪参与者 | Neuralink + 试验站点研究者 | 因颈髓损伤或 ALS 导致四肢瘫痪的成人 | 赞助方资助临床试验;没有公开常规付款路径 | Barrow、Miami、UCLH / Newcastle,通过登记和转诊进入 | 用户收益轶事和任务证明较强;没有公开转化漏斗或队列留存 |
| 医院 / 神经外科试验站点 | 学术医学中心和主要研究者 | 植入团队、研究协调员、康复人员 | 研究预算和赞助方支持 | 仅限专业神经外科中心 | 指名站点和研究者公开;未披露单站点经济性或服务水平承诺 |
| 护理者 / 居家支持层 | 患者及家庭,配合站点协调 | 植入接受者和护理者 | 试验之外不明确 | 出院后的居家支持 | 入组条件和舒适度收益可见;长期护理者负担和支持义务未披露 |
| 监管利益相关方 | FDA、MHRA、HRA/HCRW、REC、IRBs 等监管与伦理机构 | 间接受影响的研究参与者 | 公共部门监督,而非报销 | 美国 IDE / 英国早期可行性批准 | 监管闸门很明确;商业化时间和上市后义务仍不清楚 |
| 未来扩展需求登记库 | Neuralink 登记库和未来研究团队 | 截瘫、视力丧失、耳聋或失语人群 | 尚无公开付款方路径 | 目前在美国登记纳入;未来再进入试验 | 需求映射渠道已经存在;未披露登记规模或转化 |
| 未来辅助机械臂渠道 | 来自现有 PRIME 队列的交叉入组 | 已植入且需要实体任务辅助的用户 | 仍由赞助方资助 | CONVOY 可行性路径 | 用例正从光标控制外扩;但还不是独立商业渠道 |
各行把当前已植入用户、站点运营方、监管方和未来需求渠道分开,因为 Neuralink 尚未披露成熟商业客户名单或付款方基础。
[CU001, CU002, CU003, CU004, CU005, CU019]目前 Neuralink 客户旅程是一条摩擦很高的临床流程:从登记走到手术、调校、日常使用,并延伸到更广的自主任务。
该旅程图反映站点页面和官方更新描述的公开工作流;它呈现的是结构,而非实测漏斗。
[CU001, CU002, CU004, CU005, CU019, CU027]6.2 采用轨迹与具名验证
公开采用曲线在绝对规模上很小,但已经不再是假设。Barrow 在 2024 年初建立首个站点并完成首例人体植入;Neuralink 2024 年 5 月用户体验更新随后披露 Noland Arbaugh 的具体使用数据,包括工作日最多八小时研究使用、周末超过十小时个人使用,以及单周总计 69 小时。第二名参与者更新增加了另一类验证:Alex 不到五分钟就实现光标控制,几天内使用 CAD,并把植入物与 Quadstick 配对,获得更丰富的游戏控制。这些例子仍来自公司发布的文章,但它们把证据基础从静态演示视频推进到反复的居家和研究使用。 2025 年和 2026 年,新鲜度明显提升,因为医院开始发布自己的站点和患者更新。Miami 记录 RJ 成为第五名 PRIME 参与者,也是 UHealth Tower 首位植入患者;UCLH 记录 Paul 在当天或次日开始实现功能使用,之后又报告 2025 年 10 月至 12 月间七名 GB-PRIME 参与者接受植入。Neuralink 2026 年 1 月 Telepathy 更新声称已有 21 名参与者,并突出更多任务多样性,包括艺术、家庭沟通和辅助机械臂进食。合在一起,这对一项研究性服务构成了可信的类生产验证:具名站点、具名研究者、具名或化名患者、出院时点和真实用户任务。但它还不是广泛商业验证,因为分母仍很小,大多数证据仍由赞助方或站点生成,也没有公开来源披露从登记到植入的漏斗转化、活跃用户衰减或单次部署经济性。[CU006, CU007, CU008, CU009, CU010, CU011]
| 采用标记 | 数值 | 日期 | 来源类型 | 置信度 | 含义 | 缺失分母 / 限制 |
|---|---|---|---|---|---|---|
| 首个公开植入站点 | Barrow Neurological Institute | 2024-04 | 医院 + 公司 | 高 | 显示首个真实站点和首位参与者证明 | 未披露站点层面的吞吐量或等待名单 |
| 参与者 1 周使用量 | 一周 69 小时;工作日最多 8 小时,周末每天 >10 小时 | 2024-05 | 公司 | 中 | 说明有重复参与,不只是一次性演示 | 单个患者;没有队列平均值 |
| 参与者 2 激活速度 | <5 分钟开始控制光标;数天内使用 CAD 和游戏 | 2024-08 | 公司 | 中 | 暗示入门曲线可随方案迭代改善 | 仍是单名参与者,且由公司报告 |
| 第二个美国站点上线 | University of Miami / The Miami Project 入选 | 2025-01 | 医院 / 站点 | 高 | 显示地理覆盖走出 Phoenix | 除已命名中心外,没有签约站点数量 |
| Miami 植入里程碑 | 第 5 位 PRIME 参与者,也是 UHealth Tower 首例 | 2025-06 | 医院 + 站点 + 新闻 | 高 | 确认多患者、多站点运营可以复现 | 仍处于试验规模 |
| 英国首位患者功能表现 | 术后次日控制光标;居家使用并每日校准 | 2025-10 至 2025-11 | 医院 + 医院转载材料 | 高 | 显示美国之外也能当天或近乎立即产生功能性使用 | 没有长期生活质量序列 |
| 英国队列规模 | 2025 年 10-12 月,UCLH 植入 7 名 GB-PRIME 患者 | 2026-01 | 医院 | 高 | 证明站点层面积累速度快过单患者演示 | 仅一个站点的队列 |
| Telepathy 参与者数量 | 已宣布 21 名参与者 | 2026-01 | 公司 | 中 | 传递用户基数在扩大,但植入用户仍处早期 | 没有登记分母、活跃用户率或地域拆分 |
| 信号质量学习曲线 | 据称后续 20 名参与者中有 18 名在缓解措施后信号质量更高 | 2026-01 | 公司 | 中 | 暗示运营学习正在提高部署一致性 | 没有公开的同行评议耐久性数据集 |
这张表混合了参与者、站点和使用率标记,因为 Neuralink 披露的运营证明多于商业账户指标。仅由公司提供的主张维持中等置信度。
[CU006, CU007, CU008, CU010, CU011, CU012]| 指名证明实体 | 细分群体 | 部署 / 用例 | 生产 / 试点 | 结果证明 | 新鲜度 / 限制 |
|---|---|---|---|---|---|
| Noland Arbaugh @ Barrow | SCI 患者;美国首位参与者 | 在床上控制电脑、玩游戏、上网 | 试点 / 早期可行性 | 一周 69 小时;光标、国际象棋、Mario Kart、直播 | 最新详细使用指标来自 2024 年公司更新;没有独立长期留存序列 |
| Alex @ Barrow | SCI 患者;第二位参与者 | 光标控制、CAD、Counter-Strike、充电器支架设计 | 试点 / 早期可行性 | <5 分钟开始控制光标;第二天使用 CAD;未报告线体回缩 | 仅公司报告;轶事之外的生产力未发表 |
| RJ @ University of Miami 患者案例 | SCI 患者;第五位参与者 | Miami 植入后控制电脑和智能手机 | 试点 / 早期可行性 | 次日出院;UHealth Tower 首例植入;患者反馈正面 | 2025 年医院 / 站点 / 新闻证明较新,但尚无纵向随访指标 |
| Paul @ UCLH | MND 患者;英国首位参与者 | 术后次日控制光标,并在家校准 | 试点 / 早期可行性 | 术后次日使用 BCI,后来在家中每日改进使用 | 通过英国医院材料具名;没有公开每周使用小时数数据集 |
| Sebastian Gomez / GB-PRIME 队列 @ UCLH | SCI 患者加更广泛的英国队列 | 控制电脑和手机;七人站点队列 | 试点 / 早期可行性 | 患者引述加 2025 年七人手术队列 | 队列身份和个体结果只部分公开 |
| 交叉入组的 CONVOY 用户 | 现有 PRIME 参与者 | 研究性辅助机械臂控制 | 试点延伸 | 官方交叉入组路径,以及 2026 年机械臂进食高光片段 | 没有公开入组人数、完成率或独立疗效数据 |
这份列举有意保持部分覆盖而非穷尽,因为 Neuralink 不公开识别每位参与者,也不披露完整植入用户名册。
[CU006, CU007, CU008, CU010, CU012, CU013]| 指标 / 信号 | 数值 | 细分群体 | 置信度 | 说明了什么 | 尽调追问 |
|---|---|---|---|---|---|
| 参与者 1 重复使用 | 一周 69 小时;工作日研究时间最多 8 小时,周末每天 >10 小时 | Noland / 首位参与者 | 中 | 术后确实存在真实重复使用 | 需要所有植入参与者的同一指标,以及随时间衰减情况 |
| 参与者 2 立即激活 | <5 分钟开始控制光标 | Alex / 第二位参与者 | 中 | 入门可能正在变快 | 需要队列的达到独立使用所需时间中位数 |
| 出院后使用 | 次日出院后报告居家使用 | RJ 和 Paul | 高 | 站点可以很快从手术推进到早期居家使用 | 需要再入院、修正手术和支持强度数据 |
| 队列留存 | 全部已植入用户 | 低 | 没有公开毛留存或净留存曲线 | 追问 30/90/180/365 天活跃用户数 | |
| 续约 / 合同期 | 医院或付款方客户 | 低 | 公开材料没有描述商业合同 | 追问站点协议、赞助方义务和终止条款 | |
| 试验后支持保证 | 研究结束后的已植入用户 | 低 | 行业来源提示支持风险,但 Neuralink 未发布支持政策 | 追问研究后谁支付维护、取出和软件支持费用 | |
| 患者满意度 | 强轶事引述;未发布标准化调查 | 具名参与者 | 中 | 用户描述自主性提升和舒适度改善 | 追问 PROMs、护理者负担和生活质量量表 |
公开证据没有披露商业或队列留存指标,因此这些 null 是有意保留。本章把轶事性的重复使用与真正的留存经济性分开。
[CU012, CU013, CU014, CU021, CU022, CU023]Neuralink 的采用路径从登记接收走向植入后的日常使用,再在同一临床网络内扩展用例。
[CU003, CU007, CU014, CU019, CU024, CU027]公开证明质量在站点新鲜度和用户结果具体性上最强,但留存可见度和商业成熟度仍最弱。
矩阵单元格是对公开证据面的定性评估,不是临床评分输出。
[CU015, CU016, CU017, CU018, CU019, CU020]公开客户证明已从 2024 年的单患者演示,推进到 2026 年初的多站点、多患者基础设施证据。
[CU006, CU007, CU008, CU010, CU012, CU013]6.3 耐久性、集中度与采购摩擦
耐久性是最大未解客户问题,因为 Neuralink 能展示重复使用,却还不能展示成熟留存指标。没有公开 NRR、GRR、流失、合同期限或续约数据,因为公司没有披露商业装机基础。证据能显示的是植入后参与者继续互动、持续校准训练,以及用例越来越多样。与之相对,GAO 和 Inside BCI 指出了整个可植入 BCI 领域的结构性问题:研究结束且缺少报销时,试验参与者可能失去支持、维护,甚至设备访问。对 Neuralink 而言,这意味着今天最好的使用验证还没有回答,一项研究结束或扩展到赞助方资助试点之外后,谁来支付修订、软件调优、取出或长期随访。 集中度和采购摩擦也很高。公开验证集中在少数机构——Barrow、University of Miami,以及 UCLH 和 Newcastle 周边的英国站点——也集中在一小群参与者身上。入组需要手术、专门研究者、照护者可用性、监管批准和医院工作流协调,完成这些之后用户才能成为活跃日常客户。支付方采用更早期:CMS 有新技术路径,但公开 BCI 分析仍称没有专门的可植入 BCI 覆盖框架,GAO 也明确把 CMS 协调列为采用障碍。再叠加线体回缩历史、神经隐私担忧和 PCRM 驱动的安全审查带来的反向包袱,图景很清楚:Neuralink 的用户验证强于收入验证,站点动能强于支付方准备度,临床兴奋度强于商业耐久性披露。[CU021, CU022, CU023, CU024, CU025, CU026]
| 扩张驱动 | 集中度 / 摩擦 | 对客户基础的影响 | 证据 | 尽调路径 |
|---|---|---|---|---|
| Barrow 之外更多试验站点 | 当前证明集中在少数旗舰中心 | 扩张取决于增加高技能神经外科站点,不只是拉动患者需求 | Barrow、Miami 和英国来源 | 索取站点管线、入门时间和单中心吞吐量 |
| 交叉入组 CONVOY 机械臂使用 | 仍依赖已植入的 PRIME 用户 | 当前扩张是在同一用户基础内加深,而非开辟新的付款方细分 | CONVOY 启动 + Telepathy 更新 | 索取 CONVOY 入组、完成和功能结果数量 |
| 患者登记库拓宽适应证 | 登记规模和转化未披露 | 显示需求建设,但不证明获客效率 | 登记公告 + 试验页面 | 索取登记量、筛查失败率和转化为植入的比例 |
| 付款方和医院报销 | 没有公开的植入式 BCI 专用支付路径 | 即使临床兴趣升温,商业采用也可能停住 | CMS、GAO、Inside BCI | 索取编码策略、NTAP 计划和付款方沟通情况 |
| 方案和监管批准 | 每个站点都需要手术、伦理和器械批准 | 采购节奏慢,牵涉多学科且监管权重高 | FDA 概览 + UCLH 批准 | 索取站点激活清单和预期时间线 |
| 研究结束后支持 | 试验后维护义务仍不清楚 | 可能损害医院和用户的留存及声誉信任 | GAO + Frontiers + PCRM 背景 | 索取支持承诺、取出政策和数据治理承诺 |
这张表把落地后扩张路径和运营瓶颈分开;即使患者故事很强,运营瓶颈也可能让研究性装机基数长期偏小。
[CU023, CU024, CU025, CU026, CU027, CU028]6.4 图表
07风险
7.1 监管、法律与治理风险
Neuralink 本质上仍是临床阶段植入项目,不是已经上市的医疗器械公司。FDA 指引和 GAO 对 BCI 的评估都指向一条从早期可行性研究走向广泛商业化的长路:侵入式植入物属于高风险监管类别,IDE 研究必须证明收益足以覆盖受试者风险;如果设备最终上市,上市后监测、登记系统和报告义务很可能都会变得关键。公共报销不是下游细节,而是核心闸门。GAO 称 BCI 开发者与 CMS 打交道很难,CMS 自身 NTAP 框架要求产品先获得上市许可,并拿出重大临床改善证据;Inside BCI 还报道,Neuralink 已经聘请联邦游说者,明确推进「此类设备的覆盖」。 这里的反向证据格外重要,因为它已不再只是对 Musk 旗下公司的历史批评。PCRM 的投诉和记录主张、Fast Company 基于 Reuters 的 FDA 所指动物实验室缺陷报道,以及 STAT 所写治疗叙事与超人类主义话术冲突,都指向同一个承销问题:Neuralink 需要监管机构、医院伙伴和未来支付方信任一家公司,而这家公司的公共叙事可能在医疗恢复与 AI 时代增强之间来回摇摆。即便这些问题单独看都不阻断进展,叠加后也会提高延期、标签收窄、额外监测要求,或因治理质量而非单纯技术表现导致估值折价的概率。[CR001, CR002, CR003, CR004, CR005, CR006]
| 风险 / 规则 / 案件 | 司法辖区 | 状态 | 可能性 | 影响 | 缓解成熟度 | 剩余暴露 | 投资含义 | 尽调路径 |
|---|---|---|---|---|---|---|---|---|
| 侵入式 BCI 从 IDE 走向上市许可及上市后负担 | 美国 / FDA | Neuralink 仍处临床研究阶段;FDA 指南强调 IDE 收益-风险审查、高风险器械路径,以及可能的上市后监测 | 高 | 关键 | 中等——公司有活跃研究和公开监管沟通,但没有公开上市许可路径已经闭合 | 高 | 如果安全性、数据质量或标签范围不及预期,商业化时间可能实质性后延 | 索取 IDE 往来文件、计划提交类型、不良事件定义和上市后登记计划。 |
| 报销和编码路径尚未建立 | 美国 / CMS 和商业付款方 | GAO 和 Inside BCI 都把覆盖作为闸门问题;CMS 路径要求先获得上市许可,并拿出实质性临床改善证据 | 高 | 关键 | 低-中——Neuralink 已开始政策定位,但没有公开覆盖胜利 | 高 | 没有支付架构,即便器械临床表现很强,也仍然难以规模化 | 索取 CMS 预提交历史、编码策略、目标付款方名单和里程碑日历。 |
| SEC / 投资者披露和动物福利审查 | 美国 / SEC 和公共监督压力 | PCRM 称,SEC 在收到与猴子死亡披露有关的投诉后重启调查;保留来源未显示该调查已公开结案 | 中 | 高 | 低 —— 公司已有公开反驳,但缺少结案文件 | 中高 | 即便不会立刻停摆,治理折价也可能扩大 | 要求提供 SEC 往来函件、董事会监督材料,以及已核对的动物研究时间线。 |
| 神经数据隐私与神经权利缺口 | 美国和国际 | 学术和法律来源称,现有隐私框架不足以覆盖神经数据和下一代 BCI | 中高 | 高 | 低 —— 公司有隐私政策,但公共监管框架仍然碎片化 | 高 | 隐私治理薄弱,可能引发未来合规成本、诉讼和支付方犹豫 | 要求提供数据流图、留存规则、模型治理控制和事件响应预案。 |
| 雇佣与工作场所安全诉讼 | 加州 / 劳动法 | 前员工指控猴子处理不安全、遭报复及孕期歧视;保留来源未显示已有公开实体裁决 | 中 | 高 | 中低 —— 指控已公开,但整改证据较薄 | 中高 | 文化控制失灵会外溢到招聘、合规和监管信任 | 要求提供诉讼状态、和解或抗辩立场、培训记录,以及投诉后的流程调整。 |
各行按剩余敞口排序,聚焦可能在大规模上市前拖慢商业化或压缩估值的公开法律与政策问题。
[CR001, CR002, CR003, CR005, CR006, CR007]仅用本章保留的公开证据,按可能性和影响定位 Neuralink 的剩余风险。
[CR008, CR014, CR015, CR029, CR030, CR038]7.2 运营、安全与患者支持风险
运营风险不只是「植入物能否工作一次」,而是 Neuralink 能否在越来越多的站点里反复完成植入、支持、修订,并在必要时取出设备。公司更新显示了真实进展,也暴露出尚未解决的工程和服务风险。Neuralink 披露,首次 PRIME 手术后部分导线回缩,导致有效电极减少;随后靠软件改进恢复了表现。后续公司材料仍在讨论提升导线留存、降低侵入性,说明公司自己也把可靠性视为仍在推进的工作流,而非已经关闭的问题。早期平台出现这种情况并不异常,但它让耐久性和修订经济性成为核心尽调议题,而不是未来细节。 患者支持同样关键。GAO 明确警告,一些 BCI 受试者曾因试验后没有资金或医疗支持而被移除设备;Columbia 的生物伦理分析则认为,放弃支持可能让已经恢复的沟通能动性本身崩塌。Neuralink 自身更新显示了有意义的功能收益,包括机械臂使用;但本次保留的公开记录仍未量化取出设备准备金、修订义务、不良事件率,或按司法辖区划分的长期连续支持。再叠加 2024 年员工诉讼中的劳动与安全指控,以及临床前工作里 FDA 指出的动物实验室缺陷,运营图景已经清楚:Neuralink 的核心下行风险不是一个已经在人类身上证实的单一灾难事件,而是一组尚未解决的流程风险。装机基数一旦扩大,这些风险可能变得昂贵、伤害声誉且难以修复。[CR004, CR005, CR016, CR017, CR022, CR023]
| 失效模式 | 发生可能性 | 影响 | 缓释成熟度 | 剩余敞口 | 未解决缺口 |
|---|---|---|---|---|---|
| 线体回缩或电极丢失事件降低有效通道数和用户表现 | 中高 | 高 | 中 —— Neuralink 已调整算法,并在改进线体留存 | 高 | 未找到公开的修订率、硬件故障率或队列耐久性表。 |
| 长期支持、修订或取出义务超出赞助方假设 | 中高 | 关键 | 低 —— GAO 和伦理来源已指出这一问题,但公开资金或储备计划缺位 | 高 | 未找到公开的植入物取出储备、试验后延续协议,或按司法辖区划分的支持政策。 |
| 网络向国际扩张后,跨站点手术可复现性和随访复杂度上升 | 中 | 高 | 中 —— 旗舰学术站点降低短期风险,但也掩盖集中度 | 中高 | 未找到公开的逐站点吞吐量、培训节奏或并发症差异数据。 |
| 神经数据隐私或安全事件伤害信任,并触发新监管 | 中 | 高 | 低 —— 政策关注已可见,但公开技术控制披露不深 | 高 | 未找到公开的第三方保证包或神经数据专项控制审计。 |
| 不良事件透明度落后于产品迭代和站点扩张速度 | 中 | 高 | 中低 —— 公司更新披露了一些挫折,但不是完整事件台账 | 高 | 公开材料未量化严重事件、取出、修订或停用率。 |
运营风险以把试验阶段植入物扩展为服务的视角来框定;这项服务必须在多个站点、多年周期里保持安全、性能和可支持性。
[CR004, CR005, CR016, CR017, CR022, CR023]Neuralink 主要风险如何传导为商业化延迟、融资需求和估值压力。
[CR022, CR023, CR038, CR039, CR040, CR043]7.3 伙伴、站点与依赖风险
Neuralink 当前进展依赖少数外部机构,每一家都扮演闸门角色。Barrow、Miami、UHN、Cleveland Clinic Abu Dhabi 这样的旗舰医院不是可互换的分销渠道;它们是高度专业化的临床伙伴,提供外科人才、患者筛选、监管接口和植入后随访。UHN 材料同时凸显机会和集中度:Toronto Western 是加拿大独家手术站点,也是首个进行 Neuralink 手术的美国以外站点;UAE 项目则集中在 Cleveland Clinic Abu Dhabi,并与当地 Department of Health 绑定。集中有战略价值,因为它允许谨慎扩张;但也意味着任何站点级暂停、研究者流失、不良事件或特定司法辖区政策变化,都可能不成比例地拖慢入组和学习速度。 覆盖与商业化依赖同样集中。Inside BCI 的报道清楚表明,即便尚未获得上市许可,Neuralink 已经在推进 Medicare 架构;这强烈说明管理层理解,商业化要穿过报销和编码,而不只是拍更多患者视频。尽调上的含义是,伙伴依赖有多层:Neuralink 依赖医院站点执行试验,依赖监管机构延续研究并确定标签范围,依赖支付方和政策中介最终兑现收入,也依赖投资者撑过这些闸门清除前的阶段。这样的组合让公司比典型软件创业公司,甚至比进入市场路径更简单的普通医疗科技创业公司更脆弱。[CR008, CR009, CR018, CR019, CR020, CR021]
| 依赖项 | 对手方 / 守门人 | 角色 | 集中度 | 失效情景 | 严重性 | 缓释 | 剩余敞口 |
|---|---|---|---|---|---|---|---|
| 临床授权 | FDA、Health Canada、UAE 卫生主管部门、伦理机构 | 批准研究、修订案和最终上市路径 | 高 | 研究暂停、标签更窄、修订案审批变慢,或新增监测要求 | 关键 | Neuralink 已在多个司法辖区运营,并较早接触监管方 | 高 |
| 试验执行与患者随访 | Barrow、Miami、UHN、Cleveland Clinic Abu Dhabi、其他旗舰站点 | 手术、入组、监测和受试者支持 | 高 | 站点暂停或研究者流失会削弱入组、学习速度和公开验证 | 高 | 知名医院降低可信度风险,也集中专业能力 | 中高 |
| 商业报销架构 | CMS、编码机构、商业支付方、政策中介 | 把临床获益转化为支付和准入 | 高 | 设备临床前景仍好,但商业上沦为小众或由赞助方出资 | 关键 | 授权前已启动游说和早期政策工作 | 高 |
| 资本基础 | 后期投资者和未来融资方 | 资助试验、制造扩张、支持基础设施和新适应症 | 中高 | 报销和上市授权到来前,需要以更差条款补充资本 | 高 | Series E 轮争取了时间,但没有公开的自我造血模型 | 中高 |
| 国际扩张栈 | 医院系统,加上当地卫生部和物流伙伴 | 开拓新司法辖区,并在各地维持标准 | 中 | 地理扩张带来的合规负担快于运营纪律成熟 | 高 | 初期扩张借助精英机构,而不是广泛渠道伙伴 | 中高 |
这些依赖项按打断路径的直接程度排序:从有前景的试验走向可报销、可支持的产品业务。
[CR008, CR009, CR018, CR019, CR020, CR021]Neuralink 依赖核心外部机构来执行试验、推进商业化并建立信任。
[CR008, CR018, CR019, CR020, CR021, CR025]7.4 资本模型与执行风险
财务模型风险不是近期资不抵债,而是从今天的资本基础走到可复制商业业务之间,还有多少未定价的执行工作。Neuralink 2025 年 Series E 规模很大,达到 $650 million,投资人名单也很亮眼;但公司和第三方融资报道都把资金描述为扩展可及性、增加站点、打造未来设备的燃料,而不是自我维持经济性的证明。同一批公开材料一边庆祝五名活跃用户和全球试验,一边没有披露收入、烧钱速度、单位成本、毛利率或现金跑道。信息不透明迫使投资者把公司更像一个长周期风险平台来对待,而不是一个可以承销的医疗科技运营模型。 并行推进放大了执行风险。Neuralink 同时运行并扩展 Telepathy 研究,迭代硬件以提升留存并降低侵入性,准备更广泛的报销博弈,还把叙事延伸到语音和视觉项目。STAT 的报道显示,当宣传话术跑在当前试验更窄的治疗现实前面时,这种广度可能削弱可信度。员工安全诉讼和动物福利审视又带来一层人员与文化折价:即便相关主张最终没有灾难性发现,它们也提高了内部流程纪律被野心拉伸的概率。这一点重要,因为 Neuralink 最昂贵的失败模式不是错过某个技术里程碑;而是在可靠性、支持模型和披露实践还不足以支撑估值重量之前,就试图跨司法辖区、适应症和政策制度扩张。[CR010, CR016, CR017, CR025, CR026, CR027]
| 角色 / 职能 | 依赖或缺口 | 发生可能性 | 严重性 | 缓释成熟度 | 尽调路径 |
|---|---|---|---|---|---|
| 高管传播与外部定位 | 治疗性设备现实可能被增强或 AI 共生叙事模糊 | 高 | 高 | 中低 —— 治疗项目真实存在,但传播口径仍然混杂 | 要求提供董事会关于沟通治理、支付方信息和适应症优先级的材料。 |
| 工作场所安全与文化 | 员工指控指向动物操作中的流程纪律和便利安排风险 | 中高 | 高 | 低 —— 公开整改证据稀少 | 要求提供诉讼状态、HR 整改行动、EHS 审计和升级机制。 |
| 临床项目管理 | 多个站点、司法辖区和未来适应症带来并行压力 | 中高 | 高 | 中 —— 精英伙伴有帮助,但编排负担正在上升 | 要求提供项目管理架构、站点人员计划,以及新增适应症的关口标准。 |
| 披露与财务纪律 | 公开材料强调进展,但缺少烧钱、毛利和不良事件深度 | 高 | 高 | 中低 —— 融资渠道仍强,但公开透明度仍薄 | 要求提供月度现金消耗、植入单元经济、支持成本假设和事件报告节奏。 |
这里的执行风险不主要在技术人才,而在治理和运营纪律能否跟上临床、监管与叙事的同步扩张。
[CR010, CR016, CR017, CR025, CR026, CR027]7.5 缓释、监测与投资逻辑破裂点
这里确实存在缓释逻辑,但仍不成熟。Neuralink 已从单一美国可行性故事扩展为多国临床网络,公开披露部分设备挫折而不只展示成功,并且似乎在早期就投入报销策略,而不是假设临床进展自然释放收入。这些都是建设性信号。公司自己强调导线留存、降低侵入性,并与监管机构和医院站点协作,也属正面。但保留记录仍没有提供足以让外部投资者高置信度关闭主要风险的文件:详细不良事件日志、监管沟通、量化的取出与修订义务、支付方互动材料,或把进展转化为可持续商业模式的经济披露。 实际含义是,Neuralink 只有在明确监测纪律下才具备可投资性。干净的投资逻辑需要证明:可靠性在提升且没有隐藏的支持成本膨胀;站点可以倍增而治理裂缝不扩大;报销工作正从抽象游说推进到实际编码和覆盖牵引。如果严重安全或披露争议重现,如果医院或监管支持变得更有条件,或者资本需求持续复合却看不到更清晰的付款路径和上市后义务,投资逻辑就会破裂。换句话说,公司上行空间仍真实存在,但今天的估值支撑应绑定分阶段证据释放,而不是笼统相信 Musk 领导下的必然性。[CR023, CR024, CR038, CR043, CR044, CR047]
| 风险 | 可监测触发项 | 阈值 / 事件 | 行动含义 |
|---|---|---|---|
| 监管路径风险 | FDA / 司法辖区往来函件和公开研究状态 | 研究暂停、授权路径窄于预期,或新增监测义务但没有相应数据质量提升 | 暂停商业化假设,重新评估时间、成本和标签范围。 |
| 报销风险 | 覆盖架构进展 | 受试者人数增长时,仍没有清晰的编码 / CMS 策略,或支付方没有进展 | 只按由赞助方出资的采用情景处理,并大幅下调收入兑现时间。 |
| 安全与支持风险 | 不良事件、修订或取出披露 | 严重事件反复、支持退出原因不明,或持续缺少量化的延续计划 | 在认可规模化前,要求完整事件和支持资料包。 |
| 治理与劳动风险 | 诉讼、监督组织投诉和董事会层面整改结案 | 新的员工安全指控、不利披露发现,或没有可信整改证据 | 施加治理折价,并重新判断管理层能否负责任地扩张。 |
| 资本模型风险 | 融资和单元经济可见度 | 在报销里程碑前需要追加资本,且未披露烧钱或成本改善 | 估值立场转向期权价值,而不是可承销的医疗科技复利。 |
终止标准设计为可从监管行动、公司披露或私下尽调资料包中监测,而不是依赖主观情绪。
[CR008, CR009, CR022, CR023, CR024, CR028]7.6 展示材料
08估值
8.1 建议、融资背景与估值立场
Neuralink 仍足够重要,值得留在投资者观察名单上。公司在 2025 年 6 月完成规模很大的 $650 million Series E,继续吸引顶级投资人阵容,并持续把其植入平台定位为侵入式脑机接口的类别领导者。这是真信号。但估值跑得远快于公开基本面。Semafor 及后续报道把该轮定在约 $9 billion 投前,Acquinox 将其表述为接近 $9.6 billion 投后,更高的基于追踪器的老股报价则远超这个水平。这些更高标记没有任何公开收入、毛利、烧钱速度或报销披露支撑。 这个缺口决定了建议。在上一轮正式融资价位上,Neuralink 看起来具备战略重要性,但财务披露不足,因此正确结论是继续研究,而不是买入。置信度只有中等,因为技术和投资人基础令人印象深刻;但风险仍高,因为商业化仍取决于 FDA 上市许可、报销架构和持久的试验后支持。新投资者至少应要求风险投资式回报,并应把 2025 年正式融资轮,而不是嘈杂的老股标记,作为相关起始锚点。[CV001, CV002, CV003, CV004, CV006, CV008]
| 维度 | 评估 | 信心 | 决策含义 |
|---|---|---|---|
| 建议 | 当前价格继续研究;若条款重置再跟踪 | 中 | 只有拿到更深入的私有数据并保持价格纪律,才继续尽调。 |
| 风险评级 | 高 | 高 | 临床、报销、治理和支持风险都可能显著牵动价值。 |
| 估值立场 | 围绕 2025/2026 锚点偏紧 | 中 | 正式轮次是可用锚点;更高的二级市场标记缺少公开支撑。 |
| 回报目标 | 入场后总回报 >=3x | 中 | 若约 $9.6B 入场,稀释前退出价值需要接近 $28.8B。 |
| 持有期 | 长周期,约 7-10 年 | 中 | 没有公开证据指向近期流动性。 |
| 基准承销区间 | 约 $6B-$10B | 低 | 以正式 2025 轮为中心,但因缺少收入和覆盖证明而打折。 |
| 退出准备度 | 低 | 中 | 把 Neuralink 视为里程碑驱动的私有持有资产,而不是 IPO 就绪故事。 |
这张表把业务潜力和价格纪律拆开;结论不是看空产品,而是反对为不透明支付过高价格。
[CV002, CV003, CV038, CV047, CV050, CV051]结论仍是继续研究,因为战略上行、投资者支持和患者效用,被缺失的经济性、报销依赖和治理风险抵消。
[CV002, CV008, CV011, CV042, CV043, CV053]8.2 投资逻辑、反向逻辑与公开支撑测试
投资逻辑是自洽的。Neuralink 已经搭出差异化全栈系统——植入物、导线、机器人、软件和多适应症路线图——并已把它转化为患者效用和异常强的融资能力。官方 Series E 材料和独立报道都指向一家公司:在人类使用里程碑之后,它仍能吸引严肃资本;更广泛的 BCI 市场也在通过医疗、辅助,以及最终由软件定义的应用继续扩张。如果 Neuralink 能从早期验证走向可复制临床部署,公司可能成为神经科技中少数平台型资产之一。 反向逻辑是,在公开经济性出现前,价格已经被要求承担太多工作。已审阅证据仍显示这是一家临床阶段平台,没有公开收入桥,没有公开报销回款,也没有公开股权结构瀑布。与此同时,风险堆叠异常耦合:商业化前,CMS 式覆盖工作已经成为游说重点;GAO 记录的试验后支持问题可能打穿 BCI 经济性;反向来源还在持续提出治理、动物实验室和劳动问题。这意味着头部估值由可选性支撑,而不是由已披露现金生成支撑。[CV007, CV011, CV012, CV013, CV014, CV015]
| 论点 | 证据 | 什么会改变判断 |
|---|---|---|
| 正方:Neuralink 是侵入式 BCI 旗舰平台。 | 大额 Series E 融资、多适应症路线图和真实患者使用证明,让战略逻辑仍然成立。 | 临床上更优的竞争对手,或明显更弱的患者结果,会削弱品类领导地位。 |
| 正方:投资者财团仍在验证这个品类。 | 官方和独立的 2025 年来源显示,主要投资者仍在大规模资助公司。 | 如果未来轮次依赖内部支持或降价重定价,这一信号会变弱。 |
| 正方:平台期权未来可能支撑溢价。 | Telepathy、语音和未来适应症逻辑表明,一套植入架构可支持多个市场。 | 如果审批范围保持狭窄,或支持经济性仍重,平台期权应获得更小溢价。 |
| 反方:公开经济性仍缺位。 | 审阅来源均未披露收入、毛利率、烧钱、现金跑道或已实现报销。 | 经审计的商业和现金流披露会收窄估值折价。 |
| 反方:报销仍是关口依赖。 | Neuralink 已在商业化前就覆盖问题展开游说,CMS / GAO 来源显示路径复杂。 | 清晰的 FDA 申报路径,加上具体编码和支付方进展,会改善支撑。 |
| 反方:治理和执行噪音会压缩估值。 | 动物实验室处罚、SEC 相关审查和员工诉讼,提高了治理折价的可能性。 | 彻底整改、更充分披露,以及平稳的临床扩张,会移除部分压力。 |
反方论点针对估值:它认为在公开经济性出现前,当前标记已要求投资者先为未来成功买单。
[CV007, CV011, CV014, CV015, CV016, CV017]Neuralink 在市场野心和产品差异化上得分很高,但披露、报销准备度和近期退出准备度明显更低。
[CV039, CV041, CV042, CV043, CV052, CV053]8.3 可比框架与乐观 / 基准 / 悲观区间
最干净的估值框架是分阶段参照法,而不是 DCF。正式私募锚点显示,Neuralink 在 2023 年老股交易中约为 $5 billion,在 2023 年融资追踪器摘要中约为 $3.5 billion,在 2025 年正式融资中约为 $9-9.6 billion。公开医疗器械可比公司不是直接同业,但它们是有用的披露锚点,因为它们展示了当估值绑定已报告销售额时,投资者拿到的是什么:NeuroPace 在已披露数千万美元设备收入基础上,股权价值接近半个十亿美元;Inspire 以 2025 年 $912 million 收入支撑约 $2.7 billion 估值;Medtronic 和 Abbott 这样的大型多元化龙头披露深得多,市值 / 销售额代理倍数也低于未披露私募标记所暗示的水平。 这形成了较窄的基准情景和很有可选性的乐观情景。只有商业化去风险速度显著快于当前公开记录,并且报销架构开始显得可处理而不是停留在理论层面,乐观情景才成立。基准情景把 2025 年融资轮视为最强锚点,但因公开基本面仍不透明而折扣老股市场的亢奋。悲观情景则在又一轮融资先于收入证明到来,或安全、支持、政策问题扩大折价时,重置到更早锚点。[CV018, CV019, CV020, CV021, CV022, CV023]
| 情景 | 概率信号 | 估值区间(USD M) | 从约 $9.6B 入场的回报逻辑 | 关键假设 / 下行触发项 |
|---|---|---|---|---|
| 乐观 | 25% | 11000-15000 | 入场总回报约 ~1.1x-1.6x;按今天价格仍不是本垒打结果。 | 临床扩张提速,没有重大安全或政策挫折,商业化看起来更近,公众对溢价的支持持续。 |
| 基准 | 50% | 6000-10000 | 总回报约 ~0.6x-1.0x;取决于入场价和条款,轻微下行到持平。 | 2025 年正式轮次仍是最佳锚点,但收入不透明和报销依赖让折价继续存在。 |
| 悲观 | 25% | 3500-6000 | 如果另一轮融资或政策挫折重置标记,总回报约 ~0.35x-0.6x。 | 收入证明前又回到融资、安全或支持问题浮现,或 FDA / 覆盖进展停滞。 |
| 概率加权中点 | 100% | 6800-10200 | 说明当前公开证据支持继续尽调,但不支持为追踪器驱动的二级市场热情加价。 | 只有价格重置,或私有证据补上最大承销缺口,建议才会改善。 |
这些区间是情景估计,锚定正式私人市场标记、公开风险类比和阶段推进,而不是基于已披露的 DCF 级收入模型。
[CV038, CV047, CV048, CV049, CV053, CV054]| 可比公司 / 参考 | 当前指标或已披露规模 | 估值或融资锚点 | 与 Neuralink 的相关性 | 局限 |
|---|---|---|---|---|
| Neuralink 2025 正式轮 | $650M Series E 轮 | 投前估值 ~$9B;追踪器覆盖中投后估值约 ~9.6B | 当前入场讨论的主要锚点。 | 公开收入、利润率或股权结构细节未伴随该标记披露。 |
| Neuralink 基于追踪器的二级市场 | $14.9B 到约 ~40B 指示性区间 | 仅追踪器报价 | 显示私人市场想赋予多大的投机上行。 | 报价不一致且明确未经确认,因此不能支撑基准情景。 |
| Synchron 2025 Series D | 已融资 $200M;累计 $345M;10 名患者 | 私人融资,面向商业化 | 最接近的侵入式 / 相邻 BCI 融资参考,且有明确商业化表述。 | 保留来源未披露直接估值;方案侵入性低于 Neuralink。 |
| Precision 2024 Series C | 已融资 $102M;累计 $155M | 私人融资,面向临床扩张 | 显示另一家严肃 BCI 平台仍处于更小的已披露资本规模。 | 保留来源未披露直接估值。 |
| NeuroPace 上市公司 | 过去 12 个月 RNS 收入约 ~$81.7M | 市值约 ~$533M | 有真实收入和融资约束的上市植入设备现实校验。 | 适应症和产品类别不同;市值代理不是 EV。 |
| Inspire 上市公司 | 2025 年收入 $912M | 市值约 ~$2.744B | 商业化神经刺激类比,报销敏感性已可见。 | 睡眠呼吸暂停神经刺激在临床和商业上比 BCI 更成熟。 |
| Medtronic 上市公司 | FY26 收入 $36.4B | 市值约 ~$102.96B | 顶级设备披露锚点,也是多元化战略买家参考。 | 业务过于多元,不能作为直接定价同业。 |
| Abbott 上市公司 | 2025 年销售额 $44.3B | 市值约 ~$153.59B | 另一家大型上市公司披露锚点,展示已报告基本面的样子。 | 业务过于多元,不能作为直接 BCI 同业。 |
这组可比的有用结论是披露质量和阶段纪律,而不是逐一匹配同业。Neuralink 因品类雄心应有溢价,但不应因不透明而获得无限溢价。
[CV002, CV003, CV018, CV019, CV020, CV021]里程碑信心和披露质量对可辩护估值区间的影响,远大于叙事性老股传闻能够支撑的范围。
条形反映依赖里程碑的估值锚,而不是 DCF。它们是基于正式私募标记、较早锚点和阶段折价构建的情景估计。
[CV004, CV005, CV037, CV047, CV048, CV049]该区间显示:若临床、覆盖或融资里程碑滑坡,今天价格下新投资者的上行空间相对下行风险有多窄。
区间是分析师估计,绑定阶段推进和正式私募锚点。最后一项把这些数值转换为相对大约 $9.6B 入场价的总价值结果。
[CV047, CV048, CV049, CV050, CV051, CV052]8.4 进入纪律、退出准备度与投资逻辑破裂触发器
因此,进入纪律比品牌兴奋更重要。纪律严格的投资者应从上一轮正式融资价或更低水平出发,再要求公开来源无法提供的私下证据包:经审计收入和烧钱速度、完整优先股堆叠、报销规划、不良事件和支持义务,以及制造或单位经济性细节。没有这套资料包,为基于追踪器的老股标记加价,等于承销叙事动能,而不是可承销的基本面。 退出准备度仍低。已审阅来源没有给出公开 IPO 时间表;业务读起来仍像长周期临床和政策建设,而不是近期流动性故事。这把预期持有期推到七到十年区间,也让回报数学变得苛刻:即便从约 $9.6 billion 进入、只要求简单 3x 毛回报,也需要接近 $28.8 billion 的退出价值,且未计稀释。投资逻辑破裂触发器很直接:低价轮或结构上弱于 2025 年的融资轮、FDA 或覆盖进展停滞、严重支持或安全失败,或证据显示制造和报销落后于公司估值叙事。[CV038, CV039, CV040, CV041, CV045, CV046]
| 触发项 | 阈值 / 事件 | 对论点的传导 | 行动含义 |
|---|---|---|---|
| 融资重估 | 新一轮定价融资低于 2025 年正式估值,或融资高度依赖复杂条款 | 这会说明私募市场不再接受当前估值锚。 | 除非价格补偿足够,否则立场转向回避 / 只保留期权价值。 |
| FDA / 覆盖延迟 | 从临床验证走向上市授权和覆盖机制,仍看不到进展 | 收入时点被推迟,对外部资本的依赖也被拉长。 | 下调基准情景价值,并拉长持有假设。 |
| 安全或支持失灵 | 严重不良事件形成模式、取出 / 修订担忧,或试验后支持计划薄弱 | 这会削弱平台可信度,并抬高责任和声誉成本。 | 在完整事故与支持材料完成审阅前,暂停承销。 |
| 治理升级 | 执法扩大、伦理审查悬而未决,或劳动问题反复出现 | 即便技术进展延续,也会增加治理折价。 | 要求更强控制,并压低入场价格。 |
| 经济性不及预期 | 私有数据显示毛利率薄弱、支持成本沉重,或没有报销路径 | 这会打破高端平台估值叙事。 | 按公开植入物可比公司重新定价,而不是按前沿技术期权定价。 |
| 竞争替代 | 侵入性更低的对手拿到更强证据或商业化牵引 | 这会压低 Neuralink 故事里内含的稀缺性溢价。 | 下调终局倍数和战略期权价值。 |
这些触发项都能通过新融资、监管步骤、公司披露或私有尽调资料包来跟踪。
[CV011, CV012, CV013, CV014, CV015, CV016]| 主题 | 缺失证据 | 重要性 | 负责人 / 尽调路径 |
|---|---|---|---|
| 收入与现金桥 | 经审计收入、毛利率、烧钱速度、现金跑道,以及月度现金滚动表 | 缺了这些,估值就无法和已披露基本面挂钩。 | 财务团队 / 数据室;要求提供经审计报表和月度经营包。 |
| 股权结构表与优先权 | 当前股权结构表、股份类别、清算堆叠、反稀释条款,以及任何优先债权 | 如果既有资本排在新投资人之前,回报测算会发生实质变化。 | 法务尽调;审阅章程、投资人权利文件和分配瀑布模型。 |
| 报销资料包 | 编码策略、付款方路线图、CMS 沟通状态,以及成本效益叙事 | 覆盖是临床成功走向真实收入之间的闸门。 | 市场准入和政策负责人;要求提供完整报销工作计划。 |
| 安全与支持资料包 | 不良事件、修订、取出政策、试验后连续性计划和准备金 | BCI 估值会被支持义务击穿,程度不亚于设备疗效。 | 临床、质量和医学事务尽调。 |
| 制造与单位经济性 | 机器人利用率、植入成本结构、良率、服务负担和站点搭建经济性 | 支持或制造拖累过重,会吃掉平台上行空间。 | 运营尽调;要求提供 BOM、良率、人工和服务成本模型。 |
| 退出与治理就绪度 | 董事会汇报节奏、投行沟通、IPO 准备工作和控制整改 | 这能检验持有期和治理折价是否在改善。 | CEO、CFO、总法律顾问,以及董事会材料审阅。 |
这些问题按对承销影响最直接的程度排序。第 1–4 项是估值阻断点;第 5–6 项决定高估值情景是否站得住。
[CV008, CV010, CV012, CV013, CV024, CV035]免责声明
本报告是基于截至 2026-06-13 公开信息生成的 AI 辅助尽调摘要,不构成投资建议。Neuralink 是一家私有、披露较少的临床阶段公司,因此估值和运营结论仍高度依赖非公开的临床、报销、财务和治理数据;这些数据可能实质性改变判断。
证据索引
| 编号 | 陈述 | 可信度 | 来源 |
|---|---|---|---|
| CO001 | Neuralink describes itself as a brain-computer interface company developing implantable systems that connect the brain to external devices. | 高 | SO001, SO002, SO003 |
| CO002 | Neuralink's current public product stack centers on the Link implant, its robotic implantation workflow, and branded trial programs aimed first at digital-device control. | 中 | SO002, SO003, SO016 |
| CO003 | Multiple third-party sources reviewed for this run say Neuralink was founded in 2016. | 中 | SO025, SO026, SO027 |
| CO004 | TipRanks lists Neuralink's headquarters as Fremont, California, United States. | 中 | SO024 |
| CO005 | Public reporting places Neuralink's experimental footprint in both California and Texas. | 中 | SO022, SO024 |
| CO006 | The evidence set supports classifying Neuralink as a late-stage private clinical-development company rather than a commercial-scale operating business. | 中 | SO003, SO024, SO025 |
| CO007 | Sacra characterizes Neuralink's intended business model as a vertically integrated implant-plus-surgery offering sold through hospitals to patients. | 中 | SO003, SO025 |
| CO008 | Neuralink's PRIME study is recruiting people with severe paralysis with the initial goal of thought-driven cursor or keyboard control. | 高 | SO008, SO011, SO013 |
| CO009 | Neuralink announced that the FDA had approved its first in-human clinical study in May 2023. | 高 | SO012, SO013, SO019 |
| CO010 | Neuralink opened PRIME recruitment in September 2023 after securing institutional review board and hospital-site clearance. | 高 | SO008, SO013 |
| CO011 | CNBC reported that Neuralink implanted its device in a human patient for the first time in January 2024 and said the patient was recovering well. | 中 | SO014 |
| CO012 | CNBC later identified Noland Arbaugh as Neuralink's first patient and showed him moving a cursor and playing chess using the implant. | 中 | SO015 |
| CO013 | Neuralink said part of the first implant malfunctioned after several threads retracted from Arbaugh's brain, reducing effective electrodes. | 中 | SO016 |
| CO014 | CNBC reported that Neuralink's Link system records neural signals through 1,024 electrodes distributed across 64 threads. | 中 | SO016, SO021 |
| CO015 | At its November 2022 show-and-tell, Neuralink publicly pitched human trials as roughly six months away and emphasized recruiting as it moved from prototype toward product. | 中 | SO021 |
| CO016 | CNBC reported that the U.S. Department of Transportation investigated Neuralink over alleged unsafe transport of contaminated hardware from animal experiments. | 中 | SO018 |
| CO017 | A PCRM press release described public records alleging that Neuralink monkey experiments caused infections, seizures, paralysis, internal bleeding, and other suffering. | 中 | SO022 |
| CO018 | Inc. republished a Reuters report saying FDA inspectors found "objectionable conditions or practices" at Neuralink's animal testing lab in California. | 中 | SO023 |
| CO019 | Reuters reporting carried by CNBC said the FDA had earlier rejected a Neuralink human-trial application because of concerns about wire migration, battery safety, and device removal. | 中 | SO019 |
| CO020 | Neuralink officially announced a $650 million Series E financing on June 2, 2025. | 高 | SO006, SO017, SO029 |
| CO021 | TechCrunch reported days before the company announcement that Neuralink was raising about $600 million at roughly a $9 billion pre-money valuation. | 中 | SO020 |
| CO022 | Public coverage of the 2025 financing named ARK Invest, Founders Fund, Sequoia Capital, Thrive Capital, and Lightspeed Venture Partners among participating investors. | 中 | SO017, SO029 |
| CO023 | Sacra's Neuralink profile internally presents inconsistent funding and valuation figures, including a $1.20B funding header and narrative text citing approximately $1.85B of primary funding. | 低 | SO025 |
| CO024 | TipRanks estimates Neuralink had 681 employees as of June 8, 2026. | 中 | SO024 |
| CO025 | TipRanks lists Jared Birchall as CEO, Elon Musk as co-founder and co-CEO, DJ Seo as co-founder, president and COO, and Shivon Zilis as director of operations and special projects. | 中 | SO024 |
| CO026 | Fortune reporting carried by Yahoo Finance says Birchall is listed in documents as CFO among other titles, which does not cleanly match the CEO label in third-party directories. | 低 | SO026 |
| CO027 | Fortune reporting carried by Yahoo Finance says only Musk and DJ Seo remain from Neuralink's original founding group. | 中 | SO026 |
| CO028 | Neuralink's official pages reviewed in this run do not publish a board roster, so governance assessment depends on third-party directories and media profiles. | 中 | SO001, SO005, SO024, SO026 |
| CO029 | Sacra's business-model analysis says Neuralink plans to monetize through implant hardware, surgery, and hospital-mediated deployment rather than current consumer sales. | 中 | SO025 |
| CO030 | Neuralink's official site frames digital-device control as the near-term use case and speech, vision, and deeper human-AI connection as longer-term ambitions. | 高 | SO001, SO002, SO003 |
| CO031 | Coverage from 2023 and 2024 still described Neuralink as being several validation steps away from final FDA commercialization approval. | 中 | SO012, SO013, SO014 |
| CO032 | Neuralink's careers page shows active hiring across engineering, clinical, manufacturing, and operations functions. | 中 | SO004 |
| CO033 | The sources reviewed for this chapter did not publicly disclose revenue, ARR, or customer count for Neuralink. | 中 | SO001, SO003, SO005, SO024, SO025 |
| CO034 | Business Wire reported that Neuralink paid UC Davis $1.4 million between 2017 and 2020 to use the university's facilities and animals. | 中 | SO022 |
| CO035 | The Reuters story carried by Inc said the FDA viewed the 2023 animal-lab findings as serious enough for voluntary remediation but not formal enforcement action. | 中 | SO023 |
| CO036 | Neuralink reached human-trial approval only after a period marked by hazardous-transport allegations, animal-welfare criticism, and documented FDA safety questions. | 中 | SO018, SO019, SO022, SO023 |
| CO037 | The best-supported footprint today is Fremont as headquarters with operations or experimentation activity in California and Texas. | 中 | SO022, SO024 |
| CO038 | Material public gaps remain on board composition, ownership percentages, revenue scale, customer count, and a single canonical funding total. | 中 | SO001, SO003, SO024, SO025, SO026 |
| CO039 | Because the official 2025 round size is clear but third-party funding and valuation tallies disagree, diligence should use ranges rather than a single canonical figure for total raised and valuation. | 中 | SO017, SO020, SO025, SO029 |
| CO040 | The path from 2023 FDA approval to 2024 first-patient proof and 2025 large financing supports a late-stage clinical-development designation for Neuralink. | 中 | SO012, SO014, SO015, SO017, SO020, SO029 |
| CO041 | TechCrunch reported in June 2025 that Neuralink said five people with severe paralysis had already received its implants. | 中 | SO029 |
| CO042 | TechCrunch reported that Neuralink's Blindsight program received FDA breakthrough-device clearance in September 2024 while cautioning that the milestone did not prove blindness had been cured. | 中 | SO028 |
| CM001 | Barrow's PRIME study page says Neuralink is recruiting adults with tetraplegia due to spinal cord injury or ALS for an investigational brain-computer-interface implant study. | 高 | SM002, SM003 |
| CM002 | The PRIME study is framed around thought-based computer control because existing input systems often fail many people with tetraplegia, increasing dependency and reducing social engagement. | 中 | SM003 |
| CM003 | The FDA issued final guidance in 2021 for implanted brain-computer-interface devices for patients with paralysis or amputation. | 高 | SM004, SM011 |
| CM004 | FDA says most neurological devices are class II or class III and cites deep brain stimulators as class III high-risk devices. | 高 | SM004, SM019 |
| CM005 | Neuralink's commercial path therefore belongs to the regulated implantable-medical-device market rather than a general consumer-electronics market. | 高 | SM003, SM004 |
| CM006 | WHO estimates that approximately 15.4 million people were living with spinal cord injury globally in 2021. | 高 | SM010, SM012 |
| CM007 | CDC says U.S. ALS cases were close to 33,000 in 2022 and are projected to rise by more than 10% to over 36,000 by 2030. | 高 | SM013, SM014 |
| CM008 | NINDS says people with ALS eventually may become unable to use their hands and arms or to speak while usually remaining able to understand and reason. | 中 | SM015 |
| CM009 | Brown's 2026 BrainGate release says many people with paralysis find eye-gaze and other AAC systems too slow or frustrating to use. | 中 | SM016 |
| CM010 | Tobii Dynavox describes eye tracking as a high-tech AAC method that lets people with severe disabilities control a computer and speech-generating device with their eyes. | 中 | SM018 |
| CM011 | CMS nationally covers speech-generating devices as durable medical equipment for patients with severe speech impairment. | 高 | SM020, SM021 |
| CM012 | Medtronic positions DBS as an implanted therapy for Parkinson's movement symptoms and warns of surgical complications, infection, and therapy-delivery failure risks. | 中 | SM019 |
| CM013 | USC's 2026 proof-of-concept two-way brain-interface exoskeleton project says such motor-restoration systems are promising but not yet commercially available. | 中 | SM017 |
| CM014 | Precedence Research estimates the global BCI market at $3.33 billion in 2026 and $13.86 billion by 2035. | 中 | SM007 |
| CM015 | Coherent Market Insights estimates the global BCI market at $2.75 billion in 2026 and $7.14 billion by 2033. | 中 | SM008 |
| CM016 | Mordor Intelligence estimates the neurotechnology BCI market at $1.33 billion in 2026 and $2.69 billion by 2031. | 中 | SM009 |
| CM017 | Fortune Business Insights estimates the global BCI market at $295.5 million in 2026 and $960.8 million by 2034. | 中 | SM006 |
| CM018 | The Business Research Company estimates the implantable BCI market at $1.53 billion in 2026 and $2.52 billion by 2030. | 中 | SM010 |
| CM019 | Grand View Research frames invasive BCI TAM at $168.27 billion in 2025 and non-invasive BCI SOM at $397.59 million, which is a fundamentally different sizing lens from annual revenue estimates. | 低 | SM011 |
| CM020 | Reviewed BCI market estimates differ by more than an order of magnitude because publishers use incompatible market boundaries, units, and segment definitions. | 高 | SM004, SM006, SM007, SM008, SM009, SM010, SM011 |
| CM021 | Precedence says non-invasive BCI accounted for 81.86% of revenue share in 2025. | 中 | SM007 |
| CM022 | Coherent says non-invasive BCI holds 61.7% market share in 2026. | 中 | SM008 |
| CM023 | Mordor says non-invasive systems controlled 71.35% of 2025 revenue while partially invasive formats are projected to be the fastest-growing interface type at 16.35% CAGR through 2031. | 中 | SM009 |
| CM024 | Fortune says the invasive BCI segment is projected to grow at a 22.8% CAGR during the forecast period. | 中 | SM006 |
| CM025 | Because published BCI revenue is still dominated by non-invasive systems, Neuralink's implantable market is materially smaller than broad top-line BCI TAM headlines imply. | 中 | SM006, SM007, SM008, SM009 |
| CM026 | North America leads current BCI commercialization with published regional share estimates around 39.84% to 43.97% and roughly 40.92% across reviewed analyst sources. | 中 | SM006, SM007, SM008, SM009 |
| CM027 | Coherent says hospitals dominate BCI end-user share in 2026, and Mordor says hospitals and clinics held 54.66% of 2025 share while rehabilitation centers are growing quickly. | 中 | SM008, SM009 |
| CM028 | Neuralink's near-term practical buyers are trial-capable hospitals and academic centers because the current public product is investigational, surgically delivered, and workflow-intensive. | 中 | SM003, SM008, SM009 |
| CM029 | The current effective payer for implantable BCI deployment is typically study sponsorship, grants, philanthropy, or company capital rather than routine insurance reimbursement. | 中 | SM003, SM005, SM024 |
| CM030 | CMS coverage for speech-generating devices and existing DBS payment guides show that healthcare systems can reimburse communication aids and brain implants without yet covering implantable BCIs themselves. | 高 | SM020, SM021, SM022 |
| CM031 | Boston Scientific's 2026 DBS reimbursement guide lists national average Medicare physician payments such as $1,496 for CPT 61863 and $2,206 for CPT 61867, illustrating the established coding infrastructure available for conventional brain implants. | 中 | SM022 |
| CM032 | Anthem Blue Cross classifies brain-computer-interface rehabilitation devices as investigational and not medically necessary in policy DME.00052. | 高 | SM023, SM024 |
| CM033 | Inside BCI's 2026 Health Affairs-based analysis says CMS has no dedicated coding, coverage, or payment framework for implantable BCIs and that commercial coverage is almost nonexistent. | 中 | SM024 |
| CM034 | GAO says private insurers and other public programs may use CMS coverage decisions as a guide for their own BCI policies. | 高 | SM005, SM024 |
| CM035 | GAO says some clinical trial participants lost access to implanted BCIs because there were no funds or medical support after trials ended. | 高 | SM005, SM024 |
| CM036 | Inside BCI says implantable BCIs require ongoing software updates, algorithm retraining, and multidisciplinary clinical support that conventional device frameworks do not fully capture. | 中 | SM024 |
| CM037 | Brown's 2026 BrainGate release reports that two participants achieved rapid and accurate communication at home, with one reaching 22 words per minute and a 1.6% word error rate. | 中 | SM016 |
| CM038 | GAO flags uncertainty around brain-data ownership and privacy because BCIs lack a unified framework for data ownership and control. | 高 | SM005, SM024 |
| CM039 | No reviewed public source isolates a Neuralink-specific SAM or SOM that combines candidate prevalence, implant eligibility, site capacity, pricing, and reimbursement into one supportable number. | 中 | SM006, SM007, SM008, SM009, SM010, SM011 |
| CM040 | The most defensible market view is multi-lens: demand is real and broad, but near-term commercial volumes remain constrained by trial-stage operations, reimbursement, and support continuity. | 中 | SM003, SM005, SM006, SM007, SM008, SM009, SM010, SM024 |
| CM041 | Clinical adoption of implantable BCI depends on more than FDA progress because hospitals and payers still need durable outcome evidence, budget justification, and post-implant support models. | 高 | SM004, SM005, SM023, SM024 |
| CM042 | Neuralink should be evaluated against a staged adoption path that starts in research-heavy hospitals and only later expands toward routine payer-backed care if coding, coverage, and trust improve. | 中 | SM003, SM005, SM008, SM009, SM024 |
| CP001 | Neuralink competes against a layered field that includes direct implant peers, incumbent research platforms, academic internal builds, likely entrants, and non-implant substitutes. | 中 | SP029, SP030 |
| CP002 | Synchron is a direct peer because it targets paralysis-driven digital control with an implanted BCI but uses an endovascular route instead of open-skull surgery. | 高 | SP001, SP002 |
| CP003 | Precision Neuroscience is a direct peer because it is building a brain-computer interface for thought-driven control and clinical use through a less traumatic cortical approach. | 中 | SP008, SP011 |
| CP004 | Blackrock Neurotech is an incumbent platform vendor in human BCI research because its NeuroPort hardware is FDA-cleared for distribution and widely used in intracortical studies. | 高 | SP014, SP015 |
| CP005 | BrainGate functions as an internal-build academic alternative rather than a packaged commercial product. | 中 | SP017, SP018, SP020 |
| CP006 | Paradromics is a likely entrant focused on restoring communication and computer control for people with severe motor impairment. | 中 | SP021, SP022 |
| CP007 | Tobii Dynavox is the clearest non-implant substitute for communication-first users because its TD I-Series is purchasable, eye-controlled, and purpose-built for AAC. | 中 | SP024, SP025, SP026 |
| CP008 | Medtronic DBS is not a direct communication substitute, but it is the nearest incumbent analog for implanted-brain workflow, specialist distribution, and reimbursement maturity. | 中 | SP027, SP028 |
| CP009 | Synchron raised a $200 million Series D in November 2025, bringing its total funding to $345 million. | 高 | SP005, SP007 |
| CP010 | Synchron has publicly stated that 10 patients with paralysis have received Stentrode BCIs across U.S. and Australia trials. | 中 | SP007 |
| CP011 | Synchron positions Stentrode as the first endovascular BCI placed by catheter through blood vessels rather than by opening the skull. | 高 | SP001, SP007 |
| CP012 | Synchron’s first-generation Stentrode has 16 electrodes, whereas Neuralink’s N1 is described in peer coverage as a 1,024-electrode implant, implying a trade-off between lower invasiveness and lower channel density. | 中 | SP005 |
| CP013 | Synchron’s U.S. COMMAND study enrolled six people and met its primary one-year safety endpoint with no device-related deaths or permanent increases in disability. | 中 | SP006 |
| CP014 | Synchron’s active and pending study sites span multiple U.S. and Australian hospitals, including Mount Sinai, Buffalo, UT Southwestern, Mayo Clinic, Melbourne, Sydney, and Brisbane. | 中 | SP002 |
| CP015 | Precision’s Layer 7-T received FDA 510(k) clearance on 2025-03-30 as a substantially equivalent cortical electrode device. | 高 | SP009, SP010, SP011 |
| CP016 | Precision’s cleared indication is temporary implantation for up to 30 days and brain mapping, not chronic take-home consumer use. | 高 | SP009, SP011 |
| CP017 | Precision says its Layer 7 device uses a thin flexible 1,024-electrode film that can be placed through a sub-millimeter incision in a reversible procedure. | 中 | SP011 |
| CP018 | Precision’s December 2024 Series C brought its total funding to $155 million. | 中 | SP011, SP012 |
| CP019 | Precision had publicly cited at least 27 human test patients by late 2024 and Fierce later reported that 37 patients had been studied largely in short-duration neurosurgical use. | 中 | SP011, SP012 |
| CP020 | Precision’s official site frames its strategy around scale through active hospital partners and implanted clinical-study patients rather than through direct consumer availability. | 中 | SP008 |
| CP021 | Blackrock’s NeuroPort Electrode 96 has 510(k) clearance for U.S. distribution and more than 30,000 inpatient days of human data collection. | 高 | SP013, SP014 |
| CP022 | Blackrock states that the NeuroPort Electrode has recorded or stimulated for more than eight years in a single patient with zero reported serious device-related adverse events across aggregate human use. | 中 | SP014 |
| CP023 | Blackrock’s platform is used across neuroprosthetics, sensory restoration, epilepsy, and memory research, making it broader as a platform than most single-indication startups. | 中 | SP013, SP014 |
| CP024 | Blackrock’s 2024 field review said 21 research groups had implanted 67 BCI participants globally by 2023 and that 38 of those participants used NeuroPort arrays. | 中 | SP015 |
| CP025 | Tether invested $200 million for a majority stake in Blackrock in 2024 to commercialize solutions already applied to more than 40 individuals. | 中 | SP016 |
| CP026 | BrainGate2 remains a feasibility study for people with tetraplegia rather than a commercial product line. | 中 | SP018 |
| CP027 | BrainGate recruits across Massachusetts General Hospital, VA Providence, Stanford, Emory, and UC Davis, showing that the internal-build route depends on elite academic centers rather than routine community deployment. | 中 | SP018 |
| CP028 | Brown’s 2026 BrainGate update reported at-home typing at up to 22 words per minute with a 1.6% word error rate for one participant. | 中 | SP020 |
| CP029 | Paradromics positions Connexus for synthesized speech, text, and computer control for people with ALS, stroke, spinal cord injury, or other severe motor impairment. | 中 | SP021, SP022 |
| CP030 | Paradromics says Connexus uses 421 microelectrodes, exceeded 200 bits per second in preclinical testing, and is engineered for more than 10 years of stable recordings in a fully wireless system. | 中 | SP021 |
| CP031 | Paradromics says its Connect-One clinical study began in 2026 and that Connexus remains an investigational device under U.S. law. | 中 | SP021 |
| CP032 | Public pricing for chronic implantable BCIs remains undisclosed across Synchron, Precision, Blackrock/BrainGate, and Paradromics. | 中 | SP001, SP008, SP017, SP021 |
| CP033 | Tobii’s TD I-Series is a packaged speech-generating device with eye tracking, bundled AAC software, phone/web tools, and home-control capabilities. | 中 | SP024 |
| CP034 | Tobii’s funding pathway explicitly allows insurer- or program-funded access and, where necessary, direct self-purchase. | 中 | SP025 |
| CP035 | CMS covers SGD classes such as E2500 for severe speech impairment but excludes general tablets and internet service, and it requires documentation plus written order prior to delivery for specified codes. | 中 | SP026 |
| CP036 | Medtronic’s 2026 guide provides concrete physician payment anchors for DBS implantation, including $1,496 or $2,206 for first-array lead procedures and $558 or $932 for pulse generators. | 中 | SP027 |
| CP037 | Medtronic markets DBS through specialist finders, patient support, and chronic-therapy messaging, highlighting the distribution maturity absent from trial-stage BCI startups. | 中 | SP028 |
| CP038 | Implant switching costs are high because surgical team, implanted hardware, decoding software, calibration data, and clinical follow-up are bundled together. | 中 | SP001, SP008, SP014, SP018 |
| CP039 | Multi-homing is limited at the implant layer but more plausible at the output-device layer, where Synchron already links to Apple devices and SGD users can add multiple software surfaces. | 中 | SP007, SP024 |
| CP040 | Hospital and neurosurgeon access is the main channel bottleneck because leading implant players depend on named health-system partners rather than mass procurement channels. | 中 | SP002, SP008, SP018, SP021 |
| CP041 | No implanted BCI company discussed here has yet won a fully approved medical-market position for chronic use, despite decades of trials across the category. | 中 | SP015, SP030 |
| CP042 | GAO says unresolved data rights, post-trial maintenance, and CMS coverage processes remain field-wide barriers to BCI adoption. | 中 | SP029 |
| CP043 | Synchron’s lower-invasiveness is a real adoption advantage, but its lower-channel first-generation architecture leaves room for higher-bandwidth rivals if they become safe and scalable enough. | 中 | SP005, SP007 |
| CP044 | Precision’s 30-day clearance is a material regulatory lead for short-term cortical use, but it is not proof that Precision will be first to chronic home deployment. | 中 | SP009, SP011 |
| CP045 | Blackrock and BrainGate have the deepest public evidence base in the field, but their academic and custom-stack model also demonstrates how difficult productization is. | 中 | SP015, SP018, SP020 |
| CP046 | Paradromics is the sharpest speech-focused entrant risk if its 2026 clinical study validates preclinical throughput claims. | 中 | SP021 |
| CP047 | Eye-tracking AAC and Medicare-coded SGDs reduce the number of users who will accept neurosurgery when communication is the primary unmet need. | 中 | SP024, SP026 |
| CP048 | DBS incumbents already own surgeon relationships, procedural coding, and hospital budgeting norms that BCI startups must navigate or partner around. | 中 | SP027, SP028, SP029 |
| CP049 | The field is splitting by procedure class — endovascular, surface cortical, penetrating intracortical, academic custom stack, and non-implant substitute — so buyers are choosing workflow models as much as raw signal quality. | 中 | SP001, SP008, SP014, SP021, SP024 |
| CP050 | Nature’s 2024 overview underscores that implantable BCIs are advancing quickly but still face material regulatory-cost and commercialization uncertainty as countries weigh how to govern them. | 中 | SP030 |
| CI001 | Neuralink's only publicly evidenced near-term monetization path is an implanted BCI delivered through hospitals and clinical sites rather than a consumer software subscription. | 中 | SI001, SI006, SI007 |
| CI002 | Retained public materials still frame Neuralink as an investigational clinical-trial program rather than a generally available commercial product. | 中 | SI005, SI006, SI012 |
| CI003 | No retained public source discloses a Neuralink list price, realized reimbursement rate, or standard commercial contract for the implant. | 中 | SI001, SI005, SI008 |
| CI004 | Public evidence is insufficient to determine Neuralink's revenue mix across hardware, software, and support services. | 中 | SI001, SI008, SI009 |
| CI005 | Neuralink says its 2025 financing will expand patient access and future devices, implying capital is still being directed toward trial expansion and R&D rather than self-funded operations. | 高 | SI001, SI008, SI009 |
| CI006 | Neuralink disclosed a $650 million Series E round on 2025-06-02. | 高 | SI001, SI008 |
| CI007 | Neuralink had previously disclosed a $280 million Series D round in August 2023. | 中 | SI002 |
| CI008 | Taken together, the named Series D and Series E announcements establish a public floor of at least $930 million of disclosed equity raised since 2023. | 高 | SI001, SI002 |
| CI009 | As of the June 2025 financing announcement, five people with severe paralysis were publicly described as using Neuralink to control digital and physical devices with their thoughts. | 高 | SI001, SI008 |
| CI010 | CNBC reported that Neuralink was running four separate Telepathy trials as of June 2025. | 中 | SI008 |
| CI011 | Official Neuralink materials say its clinical trials span three countries and two continents, with named neurosurgical institutions in the U.S., Canada, and the UAE. | 高 | SI001, SI009 |
| CI012 | Barrow's PRIME study page recruits adults with tetraplegia or ALS into a robot-assisted implant study for computer control, underscoring a site-led clinical GTM path. | 中 | SI006 |
| CI013 | FDA's neurological-device overview says high-risk implanted neurological devices use premarket submissions tied to risk class and may require IDE-backed clinical studies before marketing. | 中 | SI012 |
| CI014 | FDA specifically notes implanted BCIs for paralysis have dedicated non-clinical and clinical guidance and encourages pre-submission engagement to determine IDE and evidence needs. | 中 | SI012 |
| CI015 | Because Neuralink's product is a surgically implanted device, its GTM motion necessarily runs through neurosurgical centers, device-review processes, and payer coding rather than self-serve conversion. | 中 | SI006, SI012, SI016 |
| CI016 | Mature implanted-brain-device workflows already rely on specific CPT, HCPCS, hospital, and programming codes. | 中 | SI016 |
| CI017 | Medtronic's 2026 DBS guide lists separate implantation, generator, and programming activities with payment references, showing that implanted-neuro economics include an ongoing service layer beyond the device itself. | 中 | SI016 |
| CI018 | Independent policy coverage says CMS has no dedicated coding, coverage, or payment framework for implantable BCIs and that support requires software updates, algorithm retraining, and multidisciplinary clinical care. | 高 | SI011, SI021 |
| CI019 | Without a coverage determination, an FDA-approved BCI could reach market without a reimbursement mechanism for implantation or post-implant support. | 中 | SI011 |
| CI020 | CMS's NTAP materials show that novel inpatient technologies need dedicated coding and a substantial-clinical-improvement case before add-on payment is considered. | 中 | SI013 |
| CI021 | MedPAC's June 2025 report shows physician-fee-schedule accuracy and outpatient cost sharing are still active policy issues, which matters for high-touch implant economics. | 中 | SI014 |
| CI022 | AAPC's 2026 Medicare rule summary says efficiency cuts and site-of-service changes create severe reimbursement pressure for facility-based procedural services. | 中 | SI015 |
| CI023 | GAO says some BCI trial participants have already had devices removed because no funds or medical support were available after the trial period. | 高 | SI010, SI021 |
| CI024 | GAO identifies CMS coverage complexity and device maintenance/support as key adoption bottlenecks for BCIs. | 中 | SI010 |
| CI025 | Post-trial maintenance, explant, and support obligations are therefore part of the likely future service-delivery cost stack even before broad reimbursement exists. | 中 | SI010, SI011 |
| CI026 | A 2026 BCI commercialization review says prior implantable neurotech failures were driven by small treatable populations, prohibitive manufacturing costs, rehab burden, and reimbursement friction. | 中 | SI020 |
| CI027 | The same review says the Argus II reached only about 350 patients and roughly $32 million of revenue before being discontinued after nearly $300 million of investments and grants. | 中 | SI020 |
| CI028 | The commercialization review says Argus II secured about $150,000 of CMS coverage yet still faced high device costs and another roughly $350,000 of implantation and rehabilitation expense beyond the device price. | 中 | SI020 |
| CI029 | The review also says procedural-partner onboarding slowed when reimbursement was unresolved and physicians lacked precedent for the new procedure. | 中 | SI020 |
| CI030 | Nature's implantable-BCI clinical-trials review flags unsupported legacy neurotechnology cases as a commercialization risk for the field. | 中 | SI019 |
| CI031 | Medtronic's 2025 10-K says Neuroscience generated $9.8 billion of net sales in fiscal 2025 and that neuromodulation net sales grew 11%, showing there is real downstream spend once reimbursement and channel infrastructure exist. | 中 | SI017, SI018 |
| CI032 | The same filing says Medtronic's 2025 cost of products sold was $11.6 billion and SG&A was $10.8 billion, with SG&A growth driven by launches and commercialization activities. | 中 | SI017 |
| CI033 | Medtronic says liquidity planning must fund R&D, property, plant, and equipment, and other operating costs, a public-comp proxy for the capital intensity of scaling implant franchises. | 中 | SI017 |
| CI034 | Neuralink's Series E page says the company is hiring engineers and operators across disciplines, consistent with ongoing fixed-cost buildout rather than a lean software-only operating model. | 中 | SI001, SI024 |
| CI035 | No retained public source discloses Neuralink's cash on hand, monthly burn, debt, or runway months. | 中 | SI001, SI008, SI009 |
| CI036 | Public traction is limited to implanted-patient counts, trial footprint, and financing milestones; the retained pack does not disclose revenue, gross margin, utilization, or commercial customer count. | 中 | SI001, SI008, SI009 |
| CI037 | The revenue-quality question therefore cannot be underwritten from public data; at best the chapter can underwrite process milestones and capital deployment intent. | 中 | SI001, SI008, SI010 |
| CI038 | Near-term monetization will depend on coverage, coding, surgeon adoption, and post-implant support workflows at hospitals, not just device efficacy. | 中 | SI011, SI013, SI016, SI020 |
| CI039 | Neuralink's business-model risk is that it combines first-of-kind implant hardware, robot-assisted surgery, and ongoing decoder support inside a payment system not yet purpose-built for BCIs. | 中 | SI010, SI011, SI012, SI020 |
| CI040 | The latest disclosed financing materially reduces immediate insolvency risk but does not remove next-round dependence if reimbursement or commercialization slips. | 中 | SI001, SI008, SI020 |
| CI041 | The biggest investor-grade diligence blockers are realized pricing, reimbursement design, post-trial support reserves, and a full cash-burn-debt bridge. | 中 | SI010, SI011, SI013, SI020 |
| CI042 | Public evidence supports a hospital-and-payer mediated B2B2C motion rather than direct-to-consumer sales because enrollment and future payment both route through institutions. | 中 | SI006, SI011, SI016 |
| CI043 | Neuralink's trial-related official pages emphasize recruitment and user-experience milestones rather than commercial availability, reinforcing that current traction is clinical rather than revenue-bearing. | 中 | SI003, SI004, SI005 |
| CI044 | Neuralink's public technology and careers pages signal platform development and hiring but still do not offer a public procurement path or pricing cue. | 低 | SI024, SI025 |
| CI045 | The retained public pack is sufficient to conclude that revenue quality is unproven, but not sufficient to estimate normalized margin, cash conversion, or customer economics. | 中 | SI001, SI010, SI011, SI020 |
| CI046 | Underwriting today therefore rests more on financing durability and reimbursement execution than on observed recurring revenue. | 中 | SI008, SI010, SI011, SI020 |
| CI047 | 2026 Medicare payment-policy changes increase the sensitivity of hospital procedural economics, which raises the commercial bar for any expensive new implant pathway. | 中 | SI014, SI015 |
| CE001 | Neuralink's current public product family spans Telepathy for device control, CONVOY for assistive robotic-arm control, and Blindsight for future visual prosthesis use cases. | 高 | SE002, SE003, SE004, SE007 |
| CE002 | Neuralink's current device-control studies target adults with major hand-use loss due to spinal cord injury or ALS rather than a broad consumer audience. | 高 | SE002, SE011 |
| CE003 | Telepathy is positioned as a way for people with paralysis to control computers, phones, and robotic limbs using their thoughts. | 高 | SE001, SE005 |
| CE004 | The current Neuralink offer remains an investigational clinical-trial product rather than a commercially available implant system. | 高 | SE002, SE010, SE011 |
| CE005 | Blindsight is designed to bypass the natural eye-to-brain pathway and create visual perception through a specialized implant connected to a camera-driven input path. | 高 | SE003, SE024 |
| CE006 | Neuralink's implant is presented as fully implantable and cosmetically invisible. | 高 | SE001, SE028 |
| CE007 | The implant uses a hermetic biocompatible enclosure meant to survive physiological conditions harsher than those in the human body. | 高 | SE001, SE019 |
| CE008 | Neuralink says the implant is powered by a small battery that is charged wirelessly from outside the body using an inductive charger. | 中 | SE001 |
| CE009 | Custom low-power chips and electronics process neural signals on-device and transmit them wirelessly to the Neuralink Application for decoding into user actions and intents. | 高 | SE001, SE016 |
| CE010 | The implant records neural activity through highly flexible ultra-thin electrode threads designed to reduce damage during and after implantation. | 高 | SE001, SE016 |
| CE011 | Neuralink's surgical robot exists because the implant threads are too fine to be inserted by the human hand. | 高 | SE001, SE016 |
| CE012 | The public technology page says the robot head includes optics and sensors from five camera systems plus optics for an OCT system. | 中 | SE001 |
| CE013 | Neuralink's published 2019 research platform described up to 3,072 electrodes across 96 threads and robotic insertion of six threads per minute. | 高 | SE015, SE016 |
| CE014 | The 2019 technical documents describe the development platform as wired for raw-bandwidth streaming while positioning future clinical derivatives as fully implantable with wireless power and telemetry. | 中 | SE016 |
| CE015 | Neuralink's January 2026 Telepathy update says the company had 21 participants at that point. | 中 | SE005 |
| CE016 | Neuralink's January 2026 Telepathy update and its year-in-review update indicate that the first participant was implanted in January 2024 and that the product remained focused on day-to-day computer control. | 高 | SE005, SE028 |
| CE017 | UCLH reported seven GB-PRIME participants implanted between October and December 2025, showing that Neuralink had expanded the site network into the UK. | 中 | SE013 |
| CE018 | The University of Miami said its first Neuralink patient was implanted in April 2025, discharged the next day, and able to control a computer and smartphone with his thoughts. | 中 | SE012 |
| CE019 | Neuralink's second-participant update says Alex began controlling a cursor in less than five minutes and exceeded his prior assistive-technology performance within hours. | 中 | SE027 |
| CE020 | Neuralink frames PRIME as a study to show that the Link is safe and useful in the daily lives of people living with paralysis. | 高 | SE027, SE028, SE011 |
| CE021 | The public deployment model is hospital-led because implantation, training, and support are routed through partner sites such as Barrow, Miami, and UCLH rather than direct consumer distribution. | 高 | SE011, SE012, SE013 |
| CE022 | CONVOY extends the same N1 implant from device control toward investigational assistive robotic-arm control and allows cross-enrollment from PRIME participants. | 高 | SE002, SE004 |
| CE023 | Blindsight received FDA Breakthrough Device Designation but remains a future U.S. trial path rather than an approved visual product. | 高 | SE003, SE007, SE024 |
| CE024 | Neuralink says it uses a Patient Registry and prior work with advocacy groups and a consumer advisory board to inform future trial and product design. | 中 | SE006 |
| CE025 | Neuralink's careers page says BCI development requires diverse engineering, scientific, and operations expertise rather than a narrow chip-design team. | 中 | SE009 |
| CE026 | Neuralink's public GitHub organization exposes at least one active Python repository called datarepo, providing a modest developer signal beyond the main marketing site. | 中 | SE014 |
| CE027 | Neuralink's public patent trail spans surgical computer vision, network-on-chip neurological data handling, and implant enclosure design, supporting a multi-layer IP strategy rather than a single-device claim. | 高 | SE017, SE018, SE019 |
| CE028 | The 2019 paper argues that Neuralink's flexible-thread plus robotic-planning approach allows custom targeting of brain regions while avoiding vasculature, which is a core technical differentiator. | 中 | SE016 |
| CE029 | In operating-model terms, Neuralink's product loop is neural recording to on-device processing to software decoding to command execution on external devices. | 高 | SE001, SE005 |
| CE030 | The current product should be treated as a managed service workflow that bundles surgery, decoder tuning, and follow-up rather than as a standalone implant SKU. | 高 | SE011, SE012, SE013, SE027 |
| CE031 | Neuralink disclosed that thread retraction affected the first participant, then said it reduced brain motion and implant-to-brain-surface gap in the second surgery and had observed no thread retraction in participant two. | 中 | SE027 |
| CE032 | Independent sources still describe long-term safety, usability, and support for implanted BCIs as unsettled despite visible clinical progress. | 高 | SE022, SE025, SE026 |
| CE033 | Neuralink's privacy policy says clinical-trial data handling is governed by protocols, informed consents, and HIPAA authorizations where applicable. | 中 | SE008 |
| CE034 | Neuralink says it may collect health, communication, and other identifying information and may share it with providers, research partners, and service providers. | 中 | SE008 |
| CE035 | Neuralink's privacy policy says it does not sell personal information or share personal information for targeted advertising purposes. | 中 | SE008 |
| CE036 | GAO warns that BCI users face unclear data ownership and that some implanted-trial participants have lost access or support after trials ended. | 中 | SE022 |
| CE037 | KU Leuven argues that invasive BCIs can erode privacy, self-determination, and consent by giving private firms unprecedented access to neural data and cognitive processes. | 中 | SE023 |
| CE038 | PCRM's obtained records say Neuralink employees conducted invasive and often deadly rhesus macaque experiments at UC Davis between 2017 and 2020. | 中 | SE020 |
| CE039 | Frontiers says long-term effects, safety, and usability remain uncertain and also criticizes Neuralink's earlier transparency around trial-registration norms. | 中 | SE026 |
| CE040 | The public roadmap broadens Neuralink from digital autonomy toward physical-task assistance and vision restoration before any large-scale commercial rollout is visible. | 高 | SE003, SE004, SE005, SE024 |
| CE041 | Public evidence does not disclose quality-system metrics such as implant manufacturing yield, sterile field-failure rates, or replacement rates for deployed human devices. | 低 | |
| CE042 | Public evidence in this run does not provide a detailed cybersecurity architecture, third-party app interface, or third-party explant-outcome dataset for Neuralink's human implants. | 低 | |
| CU001 | Neuralink's current customer surface is an investigational clinical network of implanted users and partner sites rather than a mature commercial account base. | 中 | SU005, SU008, SU009, SU013 |
| CU002 | Public PRIME screening materials target adults with tetraplegia from cervical spinal cord injury or ALS and require stable function, English communication, and caregiver support. | 高 | SU008, SU011 |
| CU003 | Neuralink's patient registry broadens future demand mapping beyond current PRIME users to include paraplegia, vision loss, deafness, and aphasia in the U.S. | 中 | SU001 |
| CU004 | Today's buying center includes Neuralink, hospital investigators, caregivers, and regulators because implantation and follow-up are organized through study protocols rather than normal product purchasing. | 中 | SU009, SU013, SU020 |
| CU005 | Public deployment proof spans Phoenix, Miami, and a UK network around UCLH and Newcastle rather than a broad national hospital rollout. | 高 | SU007, SU009, SU013, SU015 |
| CU006 | Barrow Neurological Institute publicly identified itself as the inaugural PRIME site and tied that role to the first participant's successful implant and external-device control. | 高 | SU007, SU002 |
| CU007 | University of Miami and The Miami Project were announced as the second U.S. PRIME site in January 2025. | 高 | SU009, SU011 |
| CU008 | RJ became the fifth PRIME participant and the first implanted at UHealth Tower after receiving the device in April 2025 and leaving the hospital the next day. | 高 | SU010, SU012, SU018 |
| CU009 | The UK GB-PRIME study currently has two sites: UCLH and Newcastle Hospitals NHS Foundation Trust. | 高 | SU013, SU015 |
| CU010 | UCLH reported that seven GB-PRIME patients had surgeries between October and December 2025. | 中 | SU014 |
| CU011 | Neuralink's January 2026 Telepathy update said the program had reached 21 participants. | 中 | SU006 |
| CU012 | Noland Arbaugh used the device for 69 total hours in one week, including up to 8 hours per weekday and more than 10 hours per day on weekends. | 中 | SU002 |
| CU013 | Noland's public use cases included cursor control from bed, chess, Civilization VI, livestreaming, and Mario Kart. | 中 | SU002 |
| CU014 | Alex began controlling a cursor in less than five minutes and moved into CAD and first-person-shooter gameplay within the first days of use. | 中 | SU003 |
| CU015 | Neuralink reported that mitigations used for Alex's surgery eliminated observed thread retraction in the second participant. | 中 | SU003 |
| CU016 | Paul at UCLH could move a computer cursor with his thoughts the day after surgery and later continued using the system from home. | 高 | SU013, SU015 |
| CU017 | UCLH described Sebastian Gomez as a medical student who could use a computer and mobile phone by thought and tied that story to a seven-patient 2025 surgery cohort. | 中 | SU014 |
| CU018 | The named-site proof base is freshest in 2025-2026 hospital publications from Miami and the UK, not just in 2024 company demos. | 中 | SU010, SU013, SU014, SU018 |
| CU019 | CONVOY is a feasibility extension that cross-enrolls existing PRIME participants to control an investigational assistive robotic arm. | 中 | SU004, SU006 |
| CU020 | Public proof now spans computers, smartphones, CAD, gaming, art, and robotic-arm assistance, but much of it is still narrated by Neuralink or partner sites rather than by independent longitudinal studies. | 中 | SU003, SU006, SU014, SU015 |
| CU021 | Neuralink does not publicly disclose NRR, GRR, churn, renewal, or contract-length metrics because the program is not yet a mature commercial installed base. | 低 | SU005, SU021, SU022 |
| CU022 | Repeat-usage proof exists for named participants, but it is anecdotal and participant-specific rather than a disclosed retention cohort. | 中 | SU002, SU003, SU010, SU013 |
| CU023 | GAO documented that some BCI trial participants have lost device access or support after studies ended because funding or medical support was absent. | 高 | SU023, SU022 |
| CU024 | Public evidence does not show routine payer reimbursement for implantable BCIs today, and sector analysis describes almost no commercial coverage in the United States. | 高 | SU022, SU021 |
| CU025 | Hospitals would need billing and payment mechanisms for surgery, software updates, algorithm retraining, and follow-up support that current BCI reimbursement infrastructure does not clearly provide. | 高 | SU022, SU021, SU023 |
| CU026 | Current public deployment is highly concentrated in a handful of flagship institutions and a small implanted-user base rather than a diversified hospital network. | 中 | SU007, SU009, SU013, SU014, SU006 |
| CU027 | Neuralink's near-term expansion path is mostly depth within the same cohort—better device control and cross-enrollment into robotic-arm assistance—before any broad standalone sales motion is visible. | 中 | SU003, SU004, SU006 |
| CU028 | Eligibility restrictions around age, injury stability, communication ability, and caregiver support make customer acquisition a clinical-screening problem rather than a low-friction product funnel. | 中 | SU008, SU011, SU013 |
| CU029 | Neuralink deployment is unusually regulator-dense because U.S. use depends on IDE-style clinical oversight and UK use required MHRA, HRA/HCRW, and REC approvals. | 高 | SU020, SU013, SU014 |
| CU030 | Post-trial support, explant obligations, and long-term maintenance remain material customer-durability gaps even if near-term user performance keeps improving. | 中 | SU023, SU025, SU022 |
| CU031 | UCLH said GB-PRIME participants work with clinicians, researchers, and Neuralink engineers in ongoing follow-up appointments and research sessions. | 高 | SU014, SU013 |
| CU032 | Neuralink has not publicly disclosed the registry denominator, screen-fail rate, or conversion rate from patient interest to implantation. | 低 | SU001, SU005, SU009, SU014 |
| CU033 | Public customer-proof freshness improved when site-generated 2025-2026 updates replaced a proof base that had previously relied mostly on Neuralink's own 2024 participant posts. | 中 | SU002, SU010, SU013, SU014 |
| CU034 | Thread retraction in participant 1 created a real deployment blemish that Neuralink later tried to address through surgical and algorithmic changes. | 中 | SU002, SU003, SU006 |
| CU035 | PCRM's SEC-investigation narrative keeps Neuralink's animal-safety and disclosure controversy alive as a potential adoption overhang for hospitals and payers. | 中 | SU024, SU025 |
| CU036 | Frontiers argues that long-term biological response, psychological impacts, autonomy, and neuroprivacy remain unresolved issues for invasive BCIs like Neuralink's. | 中 | SU025 |
| CU037 | Independent third-party validation exists through hospital and trade-press coverage, but independent durability measurement is still thin relative to the publicity around autonomy gains. | 中 | SU015, SU016, SU017, SU018, SU019 |
| CU038 | No public pricing, contract, or recurring-revenue disclosure allows a grounded estimate of commercial customer lifetime value or concentration economics. | 低 | |
| CR001 | FDA guidance and the neurological-device overview treat implanted BCIs as risk-based medical devices that require formal premarket pathways rather than simple software-style iteration. | 中 | SR002 |
| CR002 | FDA says IDE approval generally depends on whether potential benefits justify the risks and whether remaining risks have been appropriately minimized for study participants. | 中 | SR002 |
| CR003 | FDA’s overview says postmarket requirements for neurological devices can include mandatory reporting, surveillance studies, and device registries. | 中 | SR002 |
| CR004 | GAO says implantable BCIs are not yet on the market and that brain-surgery risks currently limit participation largely to people with severe disabilities. | 中 | SR001 |
| CR005 | GAO says some BCI clinical-trial participants have had devices removed because there were no funds or medical support available after the trial. | 中 | SR001 |
| CR006 | GAO says CMS coverage decisions are a potential key hindrance to BCI adoption and that private insurers may use CMS decisions as a guide. | 中 | SR001 |
| CR007 | CMS’s NTAP framework requires FDA marketing authorization and substantial clinical improvement evidence, showing that eventual payment depends on more than trial participation alone. | 中 | SR003 |
| CR008 | Inside BCI reports that Neuralink hired its first federal lobbyists in April 2026 and explicitly tasked them with issues related to coverage of such devices. | 中 | SR004 |
| CR009 | Inside BCI interprets the lobbying hire as pre-positioning for Medicare and broader payer reimbursement before FDA marketing authorization arrives. | 中 | SR004, SR003 |
| CR010 | STAT reports that Neuralink’s public rhetoric about machine-human symbiosis and healthy-human implantation diverges from its narrower therapeutic clinical work, and rivals think that mismatch could hurt approval and insurance payment. | 中 | SR005 |
| CR011 | PCRM says the SEC reopened an investigation into Neuralink in December 2024, according to a letter Musk shared from his attorney to the SEC chair. | 中 | SR006 |
| CR012 | PCRM’s 2023 complaint says 12 previously healthy monkeys were euthanized as a direct result of problems with Neuralink’s implant. | 中 | SR007, SR008 |
| CR013 | PCRM says UC Davis records described implant-linked chronic infections, paralysis, brain swelling, loss of coordination and balance, and depression in monkeys used by Neuralink. | 中 | SR006, SR007 |
| CR014 | Fast Company, citing Reuters reporting and an FDA response, says FDA inspectors found objectionable conditions or practices at Neuralink’s California animal lab and sought voluntary remediation rather than immediate regulatory action. | 中 | SR010 |
| CR015 | PCRM argues that Musk’s monkey-death statements may have misled investors about the safety and marketability of Neuralink’s device. | 中 | SR006, SR007, SR008 |
| CR016 | Bloomberg coverage syndicated by Yahoo says former employee Lindsay Short sued Neuralink after alleged scratches from Herpes B-carrying monkeys, inadequate protective gear, retaliation, and pregnancy-related mistreatment. | 中 | SR011 |
| CR017 | HR Dive says the complaint alleges Short was demoted after safety complaints and terminated soon after requesting pregnancy accommodations. | 中 | SR012 |
| CR018 | UHN says Health Canada approved CAN-PRIME recruitment and Toronto Western Hospital is the exclusive Canadian surgical site for the trial. | 中 | SR013 |
| CR019 | UHN says it completed the first Neuralink surgeries outside the United States in August and September 2025 and that the patients will continue with follow-up appointments and research sessions. | 中 | SR014 |
| CR020 | Economy Middle East says UAE-PRIME is the first Middle East Neuralink clinical trial and is being run with Cleveland Clinic Abu Dhabi and Abu Dhabi’s Department of Health. | 中 | SR015 |
| CR021 | The ClinicalTrials.gov UAE-PRIME record identifies one Abu Dhabi location at Cleveland Clinic Abu Dhabi, underscoring single-site concentration in the UAE program. | 中 | SR016 |
| CR022 | Neuralink’s May 2024 PRIME update says some threads retracted after the first surgery, reducing the number of effective electrodes before algorithmic changes improved performance again. | 中 | SR017 |
| CR023 | Neuralink’s January 2026 Telepathy update says 21 participants are enrolled worldwide and still frames thread retention and reduced invasiveness as active improvement targets. | 中 | SR018 |
| CR024 | The same 2026 Telepathy update says Neuralink is working closely with regulatory bodies and hospital sites and reports zero serious device-related adverse events. | 中 | SR018 |
| CR025 | Neuralink’s Series E update says five individuals with severe paralysis were using the implant in mid-2025 and that trials were running across three countries and two continents. | 中 | SR019 |
| CR026 | Neuralink’s Series E update says the company raised $650 million and listed Barrow, Miami, Toronto Western, and Cleveland Clinic Abu Dhabi among the active clinical institutions. | 中 | SR019 |
| CR027 | MedTech Dive says the 2025 round was framed around expanding patient access and listed major institutional investors plus trial sites in Phoenix, Miami, Toronto, and Abu Dhabi. | 中 | SR020 |
| CR028 | The reviewed 2025-2026 company and funding releases disclose fundraising, participants, and site expansion, but they do not publish revenue, gross margin, burn, or cash-runway figures. | 中 | SR018, SR019, SR020 |
| CR029 | Petrie-Flom says neural data can reveal thoughts, emotions, and predispositions in ways current U.S. privacy rules do not fully address. | 中 | SR021 |
| CR030 | KU Leuven argues that invasive BCIs can threaten privacy, autonomy, and democratic self-rule if private firms gain access to neural data or algorithmic influence over cognition. | 中 | SR024 |
| CR031 | The arXiv regulatory paper says frameworks built for ordinary implantable medical devices are inadequate for next-generation networked BCIs and recommends proactive ethical regulation. | 中 | SR023 |
| CR032 | The Frontiers review identifies privacy, informed consent, safety, enhancement, and animal-welfare questions as core ethical challenges around Neuralink-style BCIs. | 中 | SR025 |
| CR033 | The Columbia bioethics article argues that device abandonment or failure can destroy restored communicative agency, making maintenance and continuity ethically significant rather than optional. | 中 | SR022 |
| CR034 | GAO’s 2026 horizon report says neural implants for human augmentation could compromise user privacy and security beyond medical uses. | 中 | SR026 |
| CR035 | PCRM says it obtained UC Davis monkey records only after suing the university, highlighting continuing transparency friction around adverse preclinical evidence. | 中 | SR009 |
| CR036 | UHN, the UAE trial materials, and company funding updates show that Neuralink’s current footprint is concentrated in a small number of elite hospital partners rather than a broad site network. | 中 | SR013, SR014, SR015, SR019, SR020 |
| CR037 | The Canada, UAE, and company trial materials show Neuralink now spans at least the U.S., Canada, and UAE, expanding access but multiplying regulatory and operational interfaces. | 中 | SR013, SR014, SR015, SR016, SR019 |
| CR038 | GAO, CMS, and Inside BCI together imply that commercialization depends on both marketing authorization and reimbursement architecture, not just more trial participants. | 中 | SR001, SR003, SR004 |
| CR039 | The animal-welfare, SEC, and employee-safety allegations raise the chance that ethics controversies spill into investor, labor, and regulator trust risk simultaneously. | 中 | SR006, SR007, SR008, SR010, SR011, SR012 |
| CR040 | Thread-retraction evidence in 2024 and thread-retention work in 2026 show that device reliability remains an active engineering risk rather than a solved production variable. | 中 | SR017, SR018 |
| CR041 | Neuralink’s own updates show meaningful functional gains such as robotic-arm control, but those gains still come from small cohorts and company-led disclosure rather than broad independent outcome reporting. | 中 | SR018, SR019 |
| CR042 | STAT’s messaging critique and GAO’s augmentation warning together imply that hype beyond the therapeutic use case could complicate payer and regulator trust. | 中 | SR005, SR026 |
| CR043 | Neuralink’s funding scale and global trial expansion indicate progress, but they also underline a capital-intensive model that remains dependent on outside financing until reimbursement and market authorization exist. | 中 | SR019, SR020, SR004, SR003 |
| CR044 | FDA guidance and GAO policy options both point toward surveillance, registry, or continuity structures that will matter if Neuralink moves beyond feasibility studies. | 中 | SR001, SR002 |
| CR045 | The current partner map is unusually narrow for a company with commercialization ambitions, because the same small set of hospitals, regulators, and payers also controls the next stage of scaling. | 中 | SR004, SR013, SR014, SR016, SR019, SR020 |
| CR046 | The retained public record does not show documentary closure for the SEC matter, the California employment case, or a full remediation package around the FDA-cited animal-lab deficiencies. | 中 | SR006, SR010, SR011, SR012 |
| CR047 | The retained public record does not disclose a public explant reserve, participant-support escrow, or quantified continuity plan if a study ends or a device must be removed. | 中 | SR001, SR018, SR019, SR022 |
| CR048 | The retained public record does not quantify revision surgeries, serious adverse-event rates, discontinuation rates, or cohort retention beyond qualitative company updates. | 中 | SR017, SR018, SR019 |
| CR049 | The highest-risk transmission chain is a safety, support, or ethics controversy that slows sites and regulators, delays reimbursement, and forces more capital raising on worse terms. | 中 | SR001, SR004, SR006, SR010, SR017, SR018, SR019, SR020 |
| CR050 | If Neuralink cannot show durable reliability, transparent adverse-event and support data, and real reimbursement progress, the investment case remains venture-style optionality rather than underwritable medical-device scale. | 中 | SR001, SR003, SR004, SR018, SR019, SR020 |
| CV001 | Neuralink's latest formal financing closed in June 2025 with a $650 million Series E round. | 高 | SV003, SV004, SV005, SV006 |
| CV002 | Semafor and follow-on coverage put the 2025 raise at about a $9 billion pre-money valuation. | 高 | SV002, SV004, SV005 |
| CV003 | Acquinox characterized the Series E as roughly $656.5 million at a post-money valuation of about $9.6 billion. | 低 | SV011 |
| CV004 | Reuters-reported secondary trades in 2023 valued Neuralink at about $5 billion, but Reuters also stressed that secondary trades are an imperfect gauge of value. | 中 | SV001 |
| CV005 | Tracker-style sources place Neuralink's 2023 Series D at $280 million and around a $3.5 billion valuation, providing the other major pre-2025 anchor. | 低 | SV004, SV011 |
| CV006 | Public sources and trackers suggest Neuralink had raised roughly $1.2 billion to $1.3 billion of primary capital by the 2025 round. | 中 | SV005, SV010, SV011 |
| CV007 | Official Series E materials said the new capital would expand patient access and develop future devices. | 中 | SV006 |
| CV008 | The reviewed public record still does not disclose recognized revenue, gross margin, burn, runway, or realized reimbursement for Neuralink. | 高 | SV006, SV011, SV013, SV014 |
| CV009 | The business is still better described as a clinical-stage implant platform than as a scaled commercial medical-device company. | 中 | SV011, SV012, SV013 |
| CV010 | CMS's new-technology add-on payment process requires FDA marketing authorization, coding readiness, cost support, and substantial-clinical-improvement logic before inpatient reimbursement can advance. | 中 | SV014 |
| CV011 | Inside BCI reported that Neuralink hired federal lobbyists in 2026 with work explicitly tied to development, commercialization, and “coverage of such devices,” signaling that reimbursement is a gating value driver. | 中 | SV012 |
| CV012 | GAO said some BCI trial participants have had implants removed because no funds or medical support were available after trials. | 中 | SV013 |
| CV013 | GAO also said Medicare coverage determinations are central for BCIs and that developers can find CMS interaction challenging. | 中 | SV013 |
| CV014 | STAT reported that Neuralink's mixed rhetoric around therapy, enhancement, and AI could hinder approval and payer acceptance for the broader BCI category. | 中 | SV015 |
| CV015 | Fast Company reported that FDA urged Neuralink's animal lab to fix objectionable conditions, adding governance and process-risk overhang. | 中 | SV016 |
| CV016 | PCRM said the SEC reopened an investigation tied to its complaint about Neuralink's monkey-death disclosures, adding another adverse data point around governance. | 低 | SV019 |
| CV017 | Bloomberg/Yahoo and HR Dive reported a former staffer lawsuit involving alleged scratches from infected monkeys and pregnancy-related firing claims, extending legal and execution risk. | 中 | SV017, SV018 |
| CV018 | Synchron announced a $200 million Series D in November 2025, bringing its total funding to $345 million. | 高 | SV020, SV021 |
| CV019 | Synchron said its Stentrode platform had been implanted in 10 patients with paralysis and that the new capital would accelerate commercialization and pivotal trials. | 高 | SV020, SV021 |
| CV020 | Precision Neuroscience announced a $102 million Series C in December 2024 and said cumulative capital reached $155 million. | 中 | SV022 |
| CV021 | Precision framed itself as one of the best-funded BCI companies after Neuralink while continuing to expand clinical research partnerships. | 中 | SV022, SV023 |
| CV022 | NeuroPace's 2025 Form 10-K said trailing 12-month net RNS System revenue was $81.7 million as of December 31, 2025. | 中 | SV024 |
| CV023 | The same NeuroPace filing says a credit covenant steps down only if 2026 net RNS revenue reaches at least $90.0 million, giving a disclosed near-term commercial benchmark. | 中 | SV024 |
| CV024 | Stock Analysis showed NeuroPace at roughly $533.23 million of market cap on June 12, 2026. | 中 | SV025 |
| CV025 | Inspire reported $912.0 million of revenue for full-year 2025. | 中 | SV026 |
| CV026 | Inspire widened its 2026 revenue outlook to $950 million to $1.0 billion while warning that coding and physician-reimbursement changes were affecting near-term outcomes. | 中 | SV026 |
| CV027 | Macrotrends listed Inspire at about $2.744 billion of market cap as of January 20, 2026. | 中 | SV027 |
| CV028 | Medtronic reported FY26 revenue of $36.4 billion. | 中 | SV028 |
| CV029 | Medtronic also reported $2.751 billion of Q4 neuroscience-portfolio revenue, including neuromodulation. | 中 | SV028 |
| CV030 | CompaniesMarketCap listed Medtronic at roughly $102.96 billion of market cap in June 2026. | 中 | SV029 |
| CV031 | Abbott reported $44.3 billion of full-year 2025 sales. | 中 | SV030 |
| CV032 | CompaniesMarketCap listed Abbott at about $153.59 billion of market cap in June 2026. | 中 | SV031 |
| CV033 | Multiples.vc maintained a dedicated Medical Devices Index with current EV and growth metrics as of June 12, 2026, reinforcing that public med-device pricing is benchmarked on disclosed fundamentals. | 中 | SV032 |
| CV034 | Using disclosed market-cap and revenue references yields rough market-cap-to-sales proxies of about 2.8x for Medtronic, 3.5x for Abbott, 3.0x for Inspire, and 6.5x for NeuroPace. | 中 | SV024, SV025, SV026, SV027, SV028, SV029, SV030, SV031 |
| CV035 | Unlike those public comparables, Neuralink's 2025 price cannot be cross-checked against public sales or earnings because the company has not disclosed them. | 高 | SV006, SV011, SV024, SV026, SV028, SV030 |
| CV036 | Premier Alternatives listed Neuralink at $14.9 billion as of June 2, 2025, which is already above the formal 2025 fundraising anchor. | 低 | SV010 |
| CV037 | Acquinox described early-2026 secondary activity suggesting implied valuations approaching $40 billion, while explicitly saying the figure remained unconfirmed and only indicative. | 低 | SV011 |
| CV038 | The wide dispersion between the $9-9.6 billion formal round and higher tracker-based quotes means public evidence does not support paying a premium secondary mark as a base case. | 中 | SV002, SV010, SV011 |
| CV039 | Mid-2025 official and independent sources described only five implanted patients with severe paralysis plus multi-site trials, which is meaningful proof but still a small installed base for a near-$10 billion valuation. | 中 | SV004, SV005 |
| CV040 | No reviewed public source discloses an IPO timeline or other near-term liquidity event for Neuralink. | 中 | SV006, SV011, SV012 |
| CV041 | Neuralink's commercialization path still depends on FDA marketing authorization, payer coding and coverage, and durable post-trial support infrastructure. | 高 | SV012, SV013, SV014 |
| CV042 | The positive thesis is that Neuralink has a differentiated full-stack BCI architecture, serious investor backing, multi-indication optionality, and real patient-utility proof. | 中 | SV004, SV005, SV006, SV011 |
| CV043 | The anti-thesis is that no disclosed revenue, unresolved reimbursement, and governance or legal noise make the current price hard to underwrite on fundamentals. | 中 | SV008, SV012, SV015, SV016, SV017, SV018, SV019 |
| CV044 | Comparable late-stage BCI financings are materially smaller than Neuralink's $650 million round and do not yet validate Neuralink's much higher valuation on disclosed commercial evidence. | 中 | SV018, SV020, SV022 |
| CV045 | Inspire's 2026 coding dispute shows that even a commercial neurostimulation company can lose support when reimbursement changes, reinforcing how exposed Neuralink remains before any established payment path exists. | 中 | SV026 |
| CV046 | NeuroPace shows that a public implant company can remain below $1 billion of equity value while already generating tens of millions of actual device revenue. | 中 | SV024, SV025 |
| CV047 | The strongest formal market anchor still supports using roughly the 2025 fundraising zone, not aggressive secondary chatter, as the center of any base-case underwriting. | 中 | SV002, SV003, SV004, SV011 |
| CV048 | A reasonable bull case of about $11 billion to $15 billion would require cleaner clinical scale-up, no material safety or policy setbacks, and visible commercialization de-risking faster than peers. | 低 | SV002, SV011, SV012, SV014 |
| CV049 | A bear-case reset toward roughly $3.5 billion to $6 billion is plausible if funding needs return before revenue proof or if safety, coverage, or governance problems intensify. | 中 | SV001, SV015, SV016, SV017, SV018, SV019 |
| CV050 | A gross 3x target return from a $9.6 billion entry implies roughly $28.8 billion of exit value before dilution. | 中 | SV003, SV011 |
| CV051 | Because cap-table terms are undisclosed, any liquidation preferences or anti-dilution protections would push the required exit value above that arithmetic $28.8 billion threshold. | 中 | SV006, SV010, SV011 |
| CV052 | The public evidence set supports underwriting Neuralink as a long-duration seven-to-ten-year hold rather than a near-term liquidity candidate. | 中 | SV011, SV012, SV006 |
| CV053 | Public evidence best supports a research-more recommendation, medium confidence, high risk rating, and a stretched valuation stance at the 2025-2026 price anchors. | 中 | SV002, SV008, SV011, SV012, SV024, SV026 |
| CV054 | A disciplined new investor should start entry discussions at or below the last formal round and resist paying up for unverified secondary-market enthusiasm. | 中 | SV002, SV010, SV011 |
| CV055 | The highest-priority diligence asks are an audited financial bridge, the full preference stack, a reimbursement dossier, adverse-event and support data, and manufacturing or unit-economics detail. | 高 | SV012, SV013, SV014, SV024, SV026 |
| CV056 | The thesis breaks if a new round prices below 2025, if FDA or coverage progress stalls, or if serious safety or support failures emerge. | 中 | SV001, SV012, SV013, SV014, SV015, SV016 |
| 编号 | 出版方 | 标题 | 引文 |
|---|---|---|---|
| SO001 | Neuralink | Neuralink — Pioneering Brain Computer Interfaces | |
| SO002 | Neuralink | Technology | Neuralink | |
| SO003 | Neuralink | Clinical Trials | Neuralink | |
| SO004 | Neuralink | Careers | Neuralink | |
| SO005 | Neuralink | Updates - Neuralink | |
| SO006 | Neuralink | Neuralink raises $650 million Series E | Updates | Neuralink | |
| SO007 | Neuralink | Neuralink Raises $280M Series D | Updates | Neuralink | |
| SO008 | Neuralink | Neuralink’s First-in-Human Clinical Trial is Open for Recruitment | Updates | Neuralink | |
| SO009 | Neuralink | PRIME Study Progress Update | Updates | Neuralink | |
| SO010 | Neuralink | PRIME Study Progress Update — User Experience | Updates | Neuralink | |
| SO011 | ClinicalTrials.gov | Precise Robotically IMplanted Brain-Computer InterfacE (PRIME) | |
| SO012 | CNBC | Elon Musk’s brain implant company Neuralink announces FDA approval of in-human clinical study | Neuralink, the neurotech startup co-founded by Elon Musk, announced Thursday it has received approval from the Food and Drug Administration to conduct its first in-human clinical study. |
| SO013 | CNBC | Elon Musk’s Neuralink is recruiting patients for its first human trial | |
| SO014 | CNBC | Elon Musk’s Neuralink implants brain tech in human patient for the first time | |
| SO015 | CNBC | Neuralink shares video of patient playing chess using signals from brain implant | |
| SO016 | CNBC | Neuralink’s first in-human brain implant has experienced a problem, company says | A number of threads have retracted from Arbaugh’s brain, Neuralink said in a blog post Wednesday. |
| SO017 | CNBC | Elon Musk's Neuralink raises $650 million in fresh capital | Elon Musk’s brain tech startup Neuralink has closed a $650 million funding round. |
| SO018 | CNBC | Elon Musk’s Neuralink is under investigation for possible unsafe transport of contaminated hardware | Elon Musk’s Neuralink is being investigated by regulators for allegedly packaging and transporting contaminated hardware in an unsafe manner. |
| SO019 | CNBC / Reuters | U.S. regulators rejected Elon Musk’s bid to test brain chips in humans, citing safety risks | The agency’s major safety concerns involved the device’s lithium battery; the potential for the implant’s tiny wires to migrate to other areas of the brain; and questions over whether and how the device can be removed without damaging brain tissue. |
| SO020 | TechCrunch | Elon Musk’s Neuralink raises $600M at $9B valuation | |
| SO021 | The Verge | Elon Musk claims Neuralink is about ‘six months’ away from first human trial | |
| SO022 | Business Wire / Physicians Committee for Responsible Medicine | Ahead of Elon Musk’s “Show and Tell,” Doctors Group Reveals Disturbing Truth Behind Neuralink Monkey Experiments | Hundreds of pages of public records revealed animals suffering from sloppy experiments that resulted in chronic infections, seizures, paralysis, internal bleeding, and declining psychological health before they were killed. |
| SO023 | Inc. / Reuters | Musk's Neuralink Animal Lab Cited by FDA for "Objectionable Conditions" | An animal testing laboratory at Elon Musk’s Neuralink brain technology company was found to have “objectionable conditions or practices” by the Food and Drug Administration. |
| SO024 | TipRanks | Neuralink Leadership, Clients & Company Overview - TipRanks.com | Neuralink had 681 employees as of June 8, 2026. |
| SO025 | Sacra | Neuralink valuation, funding & news | Neuralink raised a $650M Series E in April 2026, bringing its total primary funding to approximately $1.85 billion since its founding in 2016. |
| SO026 | Yahoo Finance / Fortune | The top power players at Neuralink, Elon Musk’s brain tech startup | Only two remain of the founding group—Musk and Neuralink President Dongjin “DJ” Seo. |
| SO027 | TechCrunch | Neuralink, Elon Musk's brain implant startup, quietly raises an additional $43M | |
| SO028 | TechCrunch | Neuralink's "breakthrough device" clearance from FDA does not mean it has cured blindness | |
| SO029 | TechCrunch | Elon Musk's Neuralink closes a $650M Series E | Neuralink closed a $650 million funding round, the company announced in a blog post on Monday. |
| SM001 | Neuralink | Clinical Trials | Neuralink | |
| SM002 | ClinicalTrials.gov | ClinicalTrials.gov | |
| SM003 | Barrow Neurological Institute | PRIME Study | |
| SM004 | U.S. Food and Drug Administration | Regulatory Overview for Neurological Devices | |
| SM005 | U.S. Government Accountability Office | Brain-Computer Interfaces: Applications, Challenges, and Policy Options | |
| SM006 | Fortune Business Insights | Global Brain Computer Interface Market Size, Share & Growth, 2034 | |
| SM007 | Precedence Research | Brain Computer Interface Market Size to Hit USD 13.86 Billion by 2035 | |
| SM008 | Coherent Market Insights | Brain Computer Interface Market Size and Forecast, 2026-2033 | |
| SM009 | Mordor Intelligence | Neurotechnology Brain Computer Interface Market Size, Share & Industry Analysis, 2031 | |
| SM010 | The Business Research Company | Global Implantable Brain-Computer Interfaces Market Report 2026 | |
| SM011 | Grand View Research | Brain Computer Interface Market Size | Industry Report, 2033 | |
| SM012 | World Health Organization | Spinal cord injury | |
| SM013 | Centers for Disease Control and Prevention | Amyotrophic lateral sclerosis estimated prevalence cases from 2022 to 2030, data from the National ALS Registry | |
| SM014 | Centers for Disease Control and Prevention | National ALS Registry Dashboard | Amyotrophic Lateral Sclerosis (ALS) | |
| SM015 | National Institute of Neurological Disorders and Stroke | Amyotrophic Lateral Sclerosis (ALS) | |
| SM016 | Brown University | Brain computer interface enables rapid communication for two people with paralysis | |
| SM017 | Keck School of Medicine of USC | Researchers test a two-way brain interface with wearable robotic legs that could one day restore walking and sensation after paralysis | |
| SM018 | Tobii Dynavox | What is Eye Tracking | |
| SM019 | Medtronic | Deep Brain Stimulation for Parkinson's Disease | |
| SM020 | Centers for Medicare & Medicaid Services | NCD - Speech Generating Devices (50.1) | |
| SM021 | Centers for Medicare & Medicaid Services | Speech Generating Devices (SGD) (L33739) | |
| SM022 | Boston Scientific | Deep Brain Stimulation 2026 Reimbursement Guide | |
| SM023 | Anthem Blue Cross | Medical Policy: Brain Computer Interface Rehabilitation Devices (DME.00052) | |
| SM024 | Inside BCI | Medicare Has No Plan for Brain-Computer Interfaces | |
| SM025 | Nature Reviews Bioengineering | The state of clinical trials of implantable brain–computer interfaces | |
| SP001 | Synchron | Synchron — Still you. | Synchron is pioneering interventional brain-computer interfaces, a category it created to scale neural interfaces to serve millions of people. |
| SP002 | Synchron | Join our study — Synchron | We have two active clinical studies investigating the use of our Stentrode device to restore motor control over digital devices. |
| SP003 | Synchron | News — Synchron | |
| SP004 | ClinicalTrials.gov | Study Details | NCT05035823 | COMMAND Early Feasibility Study | |
| SP005 | MedTech Dive | Synchron raises $200M to prepare for brain computer interface launch | Synchron has raised $200 million to support commercialization of its brain computer interface platform. |
| SP006 | MedTech Dive | Synchron’s brain-computer interface tech meets safety goal | The study enrolled six people with severe chronic bilateral upper-limb paralysis unresponsive to therapy to receive the device. |
| SP007 | Business Wire | Synchron Raises $200 Million Series D to Advance Brain-Computer Interface Technology | The financing brings Synchron’s total funding to $345 million. |
| SP008 | Precision Neuroscience | Precision | The brain-computer interface that will change everything. |
| SP009 | U.S. Food and Drug Administration | 510(k) Premarket Notification K242618 | Decision Date 03/30/2025. Decision: Substantially Equivalent. |
| SP010 | U.S. Food and Drug Administration | K242618 clearance letter and summary PDF | |
| SP011 | Fierce Biotech | FDA clears Precision Neuroscience’s minimally invasive brain-computer interface implant | The agency cleared Precision’s Layer 7 interface as a temporary implant for use up to 30 days. |
| SP012 | GlobeNewswire via EIN Presswire | Precision Neuroscience Raises $102 Million to Advance AI-Powered Brain Implant | The raise brings Precision’s total funding to $155 million. |
| SP013 | Blackrock Neurotech | Utah Array, Electrodes | Products | Blackrock Neurotech | |
| SP014 | Blackrock Neurotech | NeuroPort Electrode 96, Electrodes | Products | Blackrock Neurotech | Since 2004, the NeuroPort Electrode 96 has recorded data from human subjects for an aggregate of over 30,000 days with zero reported serious adverse events related to the device. |
| SP015 | Blackrock Neurotech | The state of clinical trials of implantable brain–computer interfaces | Our NeuroPort microelectrode array has been used by 13 research groups and implanted in 38 of the 67 participants as of December 2023. |
| SP016 | Tether | Tether Takes Strategic Stake in Leading Brain-Computer-Interface Company Blackrock Neurotech | Tether’s $200 million investment will primarily fund the commercialization and roll-out of Blackrock Neurotech’s innovative medical solutions that have already been successfully applied to more than 40 individuals. |
| SP017 | BrainGate | BrainGate - Turning Thought Into Action | |
| SP018 | BrainGate | Clinical Trials - BrainGate | The purpose of the pilot clinical study of the BrainGate2 Neural Interface System is to obtain preliminary device safety information and to demonstrate the feasibility of people with tetraplegia using the investigational BrainGate system to control a computer cursor and other assistive devices with their thoughts. |
| SP019 | ClinicalTrials.gov | Study Details | NCT00912041 | BrainGate2 | |
| SP020 | Brown University | Brain computer interface enables rapid communication for two people with paralysis | One participant was able to reach a top typing speed of 110 characters or 22 words per minute, with a word error rate of 1.6%. |
| SP021 | Paradromics | Connexus | Our Connexus Brain-Computer Interface (BCI) is designed to transform neural signals into communication – synthesized speech, text, and computer control – at speeds that enable natural conversations. |
| SP022 | Paradromics | Company | |
| SP023 | ClinicalTrials.gov | Study Details | NCT07357428 | Connect-One | |
| SP024 | Tobii Dynavox | TD I-Series - eye tracking-enabled SGD | TD I-Series is a light, fast and durable speech generating device purpose-built for augmentative and alternative communication. |
| SP025 | Tobii Dynavox | How to get funding for speech generating devices | If funding is not available in your location, you can choose to purchase a solution yourself with the guidance of a local Tobii Dynavox reseller. |
| SP026 | Centers for Medicare & Medicaid Services | Article - Speech Generating Devices (SGD) - Policy Article (A52469) | Desktop, laptop, tablet, smartphone and other hand-held computers are not considered DME ... Medicare will reimburse for speech generating software only when installed on a general computing device. |
| SP027 | Medtronic | Deep brain stimulation 2026 coding and payment guide | Lead implantation ... $1,496 or $2,206 for first array; generator implantation ... $558 or $932. |
| SP028 | Medtronic | Deep Brain Stimulation for Parkinson's Disease | Use our online directory to search for Parkinson’s specialists by ZIP code. |
| SP029 | U.S. Government Accountability Office | Brain-Computer Interfaces: Applications, Challenges, and Policy Options | Some clinical trial participants have had a BCI removed because there were no funds or medical support provided after the trial. |
| SP030 | Nature | United States sets the pace for implantable brain–computer interfaces | Countries are weighing the costs and benefits of how they regulate it. |
| SI001 | Neuralink | Neuralink raises $650 million Series E | We've closed our Series E funding round of $650 million... Five individuals with severe paralysis are now using Neuralink... We launched clinical trials... spanning three countries and two continents. |
| SI002 | Neuralink | Neuralink Raises $280M Series D | We're happy to announce our $280M Series D round led by Founders Fund. |
| SI003 | Neuralink | Neuralink’s First-in-Human Clinical Trial is Open for Recruitment | |
| SI004 | Neuralink | PRIME Study Progress Update — User Experience | |
| SI005 | Neuralink | Clinical Trials | |
| SI006 | Barrow Neurological Institute | PRIME Study | Participants will go through a robot-assisted surgery to implant a device designed to enable them to control a computer with their thoughts. |
| SI007 | ClinicalTrials.gov | NCT06429735 PRIME Study | |
| SI008 | CNBC | Elon Musk's Neuralink raises $650 million in fresh capital | Elon Musk’s brain tech startup Neuralink has closed a $650 million funding round... As of Monday, five patients have been implanted... Neuralink is currently carrying out four separate clinical trials around its Telepathy system. |
| SI009 | PharmaPhorum | Neuralink raises $650m to widen use of brain implants | The investment will allow it to expand patient access... and expand its headcount. |
| SI010 | U.S. Government Accountability Office | Brain-Computer Interfaces: Applications, Challenges, and Policy Options | Some clinical trial participants have had a BCI removed because there were no funds or medical support provided after the trial. |
| SI011 | Inside BCI | Medicare Has No Plan for Brain-Computer Interfaces | CMS has no dedicated coding, coverage, or payment framework for implantable BCIs... brain-computer interfaces require ongoing software updates, algorithm retraining, and multidisciplinary clinical support. |
| SI012 | U.S. Food and Drug Administration | Regulatory Overview for Neurological Devices | Once the study is approved, an Investigational Device Exemption (IDE) submission allows a device that is not approved or cleared for market... to be used in a clinical study to collect safety and effectiveness data to support a marketing application. |
| SI013 | Centers for Medicare & Medicaid Services | New Medical Services and New Technologies | New medical services and technologies used in the inpatient setting may be eligible to apply for an add-on payment known as the new technology add-on payment (NTAP). |
| SI014 | MedPAC | June 2025 Report to the Congress: Medicare and the Health Care Delivery System | The Commission recommends replacing the current-law updates to fee-for-service Medicare’s physician fee schedule with an annual update based on a portion of the growth in inflation. |
| SI015 | AAPC | Get Ready for 2026 Medicare Reimbursement Changes | The combined effect of the efficiency adjustment and PE redistribution will create severe reimbursement pressure for facility-based procedural services. |
| SI016 | Medtronic | Deep brain stimulation 2026 coding and payment guide | What’s inside: Physician coding and payment; HCPCS II device codes; Hospital outpatient coding and payment; Hospital inpatient coding and payment. |
| SI017 | Securities and Exchange Commission | Medtronic plc 2025 Form 10-K | Cost of products sold for fiscal year 2025 was $11.6 billion... Selling, general, and administrative expense for fiscal year 2025 was $10.8 billion... The increase... is primarily due to new product launches and commercialization activities. |
| SI018 | Medtronic Investor Relations | SEC Filings | |
| SI019 | Nature Reviews Bioengineering | The state of clinical trials of implantable brain–computer interfaces | Their Bionic Eyes are Now Obsolete and Unsupported... Abandoned: the human cost of neurotechnology failure. |
| SI020 | Journal of NeuroEngineering and Rehabilitation / PubMed Central | Brain-computer interface commercialization | Argus II struggled commercially due to an insufficient patient population, prohibitive manufacturing costs, and high procedural and postoperative rehabilitation costs. |
| SI021 | STAT | These challenges could delay brain-computer interfaces from becoming a reality for patients | The field will need to address... providing them with support after implantation, and working with CMS to determine coding, coverage, and payment. |
| SI022 | Reuters | Musk's Neuralink raises $650 million in latest funding as clinical trials begin | |
| SI023 | Federal Register | CY 2026 Payment Policies under the Physician Fee Schedule and Other Changes | |
| SI024 | Neuralink | Careers | |
| SI025 | Neuralink | Technology | |
| SI026 | Nevro | Financial Information - USA | |
| SE001 | Neuralink | Technology | Neuralink | Our brain-computer interface is fully implantable, cosmetically invisible... Advanced, custom, low-power chips and electronics process neural signals, transmitting them wirelessly to the Neuralink Application. |
| SE002 | Neuralink | Device Control | Neuralink | The N1 Implant... is an investigational brain-computer interface designed to help people with paralysis control external devices using only their thoughts. |
| SE003 | Neuralink | Visual Prosthesis | Neuralink | Bypassing the natural eye-to-brain pathway, Blindsight uses a video camera to capture images... and wirelessly transmits them to the brain via our specialized implant. |
| SE004 | Neuralink | CONVOY Study Launch | Updates | Neuralink | We’re excited to announce the approval and launch of a new feasibility trial, CONVOY, to extend BCI control using the N1 Implant to an investigational assistive robotic arm. |
| SE005 | Neuralink | Two Years of Telepathy | Updates | Neuralink | Our first product, Telepathy, aims to enable people with paralysis to directly control computers, phones, and robotic limbs using their thoughts alone. |
| SE006 | Neuralink | Announcing Neuralink's Patient Registry | Updates | Neuralink | We have worked with patient advocacy groups and a consumer advisory board... to guide product development and community engagement. |
| SE007 | Neuralink | Neuralink Receives Breakthrough Device Designation for Blindsight | Updates | Neuralink | We have received Breakthrough Device Designation from the FDA for Blindsight, for individuals with vision impairment. |
| SE008 | Neuralink | Privacy Policy | Neuralink | Our privacy practices in connection with clinical trials are governed by applicable clinical trial protocol(s) and related informed consents and, where applicable, HIPAA authorization forms. |
| SE009 | Neuralink | Careers | Neuralink | Developing brain-computer interfaces is an interdisciplinary challenge. We are looking to hire a wide range of people with diverse engineering, scientific, and operations expertise. |
| SE010 | ClinicalTrials.gov | NCT06429735 PRIME Study | ClinicalTrials.gov listing for the Precise Robotically Implanted Brain-Computer Interface (PRIME) study. |
| SE011 | Barrow Neurological Institute | PRIME Study | The Neuralink implant is designed to establish a wireless, digital link between the brain and computers... This study aims to validate the safety and usefulness of the Neuralink implant and surgical robot. |
| SE012 | University of Miami Miller School of Medicine | Paralyzed Veteran Surgically Implanted with Neuralink Device at The Miami Project to Cure Paralysis | RJ successfully received his implant in April 2025 and was discharged from the hospital the day after his surgery. |
| SE013 | UCLH Biomedical Research Centre | Seven patients now participating in Neuralink trial | There are now seven patients participating in the GB-PRIME study... evaluating the safety and functionality of the BCI. |
| SE014 | GitHub | Neuralink · GitHub | Developing high bandwidth brain-machine interfaces to connect humans and machines. |
| SE015 | PubMed / Journal of Medical Internet Research | An Integrated Brain-Machine Interface Platform With Thousands of Channels - PubMed | We have built arrays of small and flexible electrode 'threads,' with as many as 3072 electrodes per array distributed across 96 threads. |
| SE016 | PubMed Central | An Integrated Brain-Machine Interface Platform With Thousands of Channels | The robot features an autoinsertion mode... the surgeon retains full control... future clinical devices... will be fully implantable... with wireless power transmission and data telemetry through the skin. |
| SE017 | Google Patents | US20200085508A1 - Computer vision techniques | Computer vision techniques. |
| SE018 | Google Patents | US20190286592A1 - Network-on-chip for neurological data | Network-on-chip for neurological data. |
| SE019 | Google Patents | US20230148968A1 - Polymer enclosure with wire passthrough for implantable device | Polymer enclosure with wire passthrough for implantable device. |
| SE020 | Physicians Committee for Responsible Medicine | Original Records of Each Monkey Used by Neuralink at UC Davis | Between 2017 and 2020, Neuralink employees conducted invasive, often deadly experiments on rhesus macaque monkeys at the University of California, Davis. |
| SE021 | Physicians Committee for Responsible Medicine | Neuralink’s Cruel Brain Experiments | Neuralink’s cruel brain experiments. |
| SE022 | U.S. Government Accountability Office | Brain-Computer Interfaces: Applications, Challenges, and Policy Options | Experts identified several challenges... uncertainties in data ownership and control... [and] potential loss of access or support. |
| SE023 | KU Leuven AI Summer School | Neural Privacy and Democratic Self-Rule: Governance Challenges of Invasive Brain-Computer Interfaces | BCIs mark a novel governance challenge where private firms gain unprecedented access to neural data, potentially shifting influence over cognitive processes from individuals to commercial entities. |
| SE024 | MedicalDeviceNetwork | Neuralink obtains breakthrough designation from FDA for Blindsight device | The visual resolution will initially be basic, similar to early video game graphics. |
| SE025 | MIT Technology Review | Brain-computer interfaces face a critical test | About 25 clinical trials of BCI implants are currently underway. |
| SE026 | Frontiers | Neuralink’s brain-computer interfaces: medical innovations and ethical challenges | The long-term effects, safety, and usability of these devices remain uncertain... the company's initial lack of transparency... has drawn criticism from the scientific community. |
| SE027 | Neuralink | PRIME Study Progress Update — Second Participant | Updates | Neuralink | To reduce the probability of thread retraction in our second participant, we implemented a number of mitigations, including reducing brain motion during the surgery and reducing the gap between the implant and the surface of the brain. |
| SE028 | Neuralink | A Year of Telepathy | Updates | Neuralink | The implant, or Link, is our fully implantable, cosmetically-invisible, wireless brain-computer interface (BCI) designed to restore autonomy to people with paralysis. |
| SU001 | Neuralink | Announcing Neuralink's Patient Registry | We've launched a Patient Registry to learn more about individuals who may be interested in enrolling in future Neuralink clinical trials. |
| SU002 | Neuralink | PRIME Study Progress Update — User Experience | On weekdays, Noland contributes to research sessions for up to 8 hours per day. On weekends, personal use and recreation can exceed 10 hours per day. |
| SU003 | Neuralink | PRIME Study Progress Update — Second Participant | From the first moment Alex connected his Link to his computer, it took less than 5 minutes for him to start controlling a cursor with his mind. |
| SU004 | Neuralink | CONVOY Study Launch | The CONVOY Study will enable cross-enrolling participants from the ongoing PRIME Study. |
| SU005 | Neuralink | Clinical Trials | Connect with us and learn more about Neuralink clinical trials. |
| SU006 | Neuralink | Two Years of Telepathy | Announcing 21 Neuralink participants. |
| SU007 | Barrow Neurological Institute | PRIME Study Site Announcement | The first participant in Neuralink's PRIME Study recently demonstrated the ability to use brain activity to command an external device. |
| SU008 | Barrow Neurological Institute | PRIME Study | If you have tetraplegia due to spinal cord injury or ALS and want to explore new ways of controlling your computer, we invite you to participate in Neuralink's clinical trial. |
| SU009 | University of Miami Miller School of Medicine | The Miami Project to Cure Paralysis at the University of Miami Miller School of Medicine Selected as U.S. Site for Neuralink Clinical Trial | The Miami Project to Cure Paralysis and the Department of Neurological Surgery at the University of Miami Miller School of Medicine have been selected to become the second U.S.-based site for Neuralink's PRIME Study. |
| SU010 | University of Miami Miller School of Medicine | Paralyzed Veteran Surgically Implanted with Neuralink Device at The Miami Project to Cure Paralysis | RJ successfully received his implant in April 2025 and was discharged from the hospital the day after his surgery. |
| SU011 | The Miami Project to Cure Paralysis | The Miami Project to Cure Paralysis at University of Miami Miller School of Medicine Selected as a U.S. Site for Neuralink Clinical Trial | For the PRIME Study (NCT06429735), Neuralink is specifically looking for patients who have limited or no ability to use both hands due to cervical spinal cord injury or ALS. |
| SU012 | The Miami Project to Cure Paralysis | Paralyzed Veteran Implanted with Neuralink Device by Miami Project Doctors at University of Miami | RJ, who is paralyzed due to a spinal cord injury sustained from a motorcycle accident, is the fifth participant in Neuralink's PRIME Study and the first to be implanted at The Miami Project. |
| SU013 | UCLH Biomedical Research Centre | First UK patient uses thought to control computer hours after Neuralink implant | It was remarkable to see Paul using his brain-computer interface on the very first day after surgery; he is now using it in his own home and working hard every day to improve his calibration and control. |
| SU014 | University College London Hospitals NHS Foundation Trust | Seven GB-PRIME patients now participating in Neuralink trial | There are now seven patients participating in the GB-PRIME study involving Neuralink's brain-computer interface technology. |
| SU015 | Mirage News | UK Patient Controls Computer Post-Neuralink Implant | The study currently has two sites: University College London Hospitals NHS Foundation Trust (UCLH) and Newcastle Hospitals NHS Foundation Trust. |
| SU016 | HTN Health Tech News | UCLH shares details of successful brain-computer interface innovation | According to the trust, Paul is the first UK patient to undergo this procedure. |
| SU017 | Refresh Miami | 1st patient gets Elon Musk's Neuralink brain chip at the University of Miami, a clinical trial site | RJ received his implant at UHealth Tower in April 2025 and was discharged from the hospital the day after his surgery. |
| SU018 | Becker's Hospital Review | University of Miami hospital implants brain chip from Elon Musk's Neuralink | Patient 'RJ' ... became the fifth patient to receive the device and the first at UHealth. |
| SU019 | Newswise | Paralyzed Veteran Surgically Implanted with Neuralink Device at The Miami Project to Cure Paralysis at the University of Miami Miller School of Medicine | A paralyzed United States military veteran has been implanted with the groundbreaking Neuralink brain-computer device by surgeons at The Miami Project to Cure Paralysis. |
| SU020 | U.S. Food and Drug Administration | Regulatory Overview for Neurological Devices | Once the study is approved, an Investigational Device Exemption submission allows a device ... to be used in a clinical study to collect safety and effectiveness data to support a marketing application. |
| SU021 | Centers for Medicare & Medicaid Services | New Medical Services and New Technologies | To qualify for a new technology add-on payment, the technology must represent an advance that substantially improves ... the diagnosis or treatment of Medicare beneficiaries. |
| SU022 | Inside BCI | Medicare Has No Plan for Brain-Computer Interfaces | As things stand, there is almost no commercial coverage for BCIs in the United States. |
| SU023 | U.S. Government Accountability Office | Brain-Computer Interfaces: Applications, Challenges, and Policy Options | Some clinical trial participants have had a BCI removed because there were no funds or medical support provided after the trial. |
| SU024 | Physicians Committee for Responsible Medicine | SEC Reopens Investigation Into Elon Musk's Neuralink Likely Launched by Medical Ethics Group's Complaint About Monkey Deaths | Public records show Musk may have misled investors about device's safety. |
| SU025 | Frontiers | Neuralink's brain-computer interfaces: medical innovations and ethical challenges | The long-term safety of invasive BCIs, especially in terms of biological response and psychological impacts, remains largely unknown. |
| SR001 | U.S. Government Accountability Office | GAO-25-106952; Brain-Computer Interfaces: Applications, Challenges, and Policy Options | Some clinical trial participants have had a BCI removed because there were no funds or medical support provided after the trial. |
| SR002 | U.S. Food and Drug Administration | Regulatory Overview for Neurological Devices | The FDA will generally approve the study if the potential benefits justify the risks, and if remaining risks have been appropriately minimized. |
| SR003 | Centers for Medicare & Medicaid Services | New Medical Services and New Technologies | As finalized in the FY 2024 IPPS final rule, applicants for NTAP must receive FDA marketing authorization by May 1 of the year prior to the beginning of the fiscal year for which the application is being considered. |
| SR004 | Inside BCI | Neuralink hires its first federal lobbyists to open the brain-computer interface coverage account | The scope is described as issues related to the development and commercialization of brain-computer interfaces and matters related to coverage of such devices. |
| SR005 | STAT | What does Neuralink want — to help people with paralysis, or prepare for a war with AI? | This conflicting messaging from a company perceived as the emerging field’s leader could hinder the ability of startups developing brain-computer interfaces to gain approval for them as medical devices and be paid for by health insurers. |
| SR006 | Physicians Committee for Responsible Medicine | SEC Reopens Investigation Into Elon Musk’s Neuralink Likely Launched by Medical Ethics Group’s Complaint About Monkey Deaths | This week, the Commission has also reopened an investigation into Neuralink. |
| SR007 | Physicians Committee for Responsible Medicine | Physicians Group Asks SEC to Investigate Elon Musk for Securities Fraud Stemming from False Statements About Neuralink Monkey Deaths | Public Records Show 12 Monkeys Died as a Direct Result of Neuralink Implants. |
| SR008 | Physicians Committee for Responsible Medicine | Physicians Group to SEC: Elon Musk’s DealBook Summit Claims About Dead Monkeys May Be Securities Fraud | Public Records Show 12 Monkeys Died as a Direct Result of Neuralink Implants, but Musk Said Animals Had terminal case of cancer or something like that. |
| SR009 | Physicians Committee for Responsible Medicine | Original Records of Each Monkey Used by Neuralink at UC Davis | Between 2017 and 2020, Neuralink employees conducted invasive, often deadly experiments on rhesus macaque monkeys at the University of California, Davis. |
| SR010 | Fast Company | FDA urged Musk’s Neuralink animal lab to fix its objectionable conditions | The FDA told Blumenauer the issues found at Neuralink warranted voluntary remediation measures by the company. |
| SR011 | Yahoo Finance / Bloomberg | Former Neuralink Staffer Sues After Scratches From Herpes Monkey | Short said she was working with monkeys that carried the Herpes B virus when she was scratched through a glove. |
| SR012 | HR Dive | Ex-Neuralink staffer claims she was scratched by infected monkeys, fired over her pregnancy | The plaintiff also claimed that Neuralink confronted her about legitimate time off requests to care for family... she alleged she was terminated soon after. |
| SR013 | University Health Network | Neuralink for Patients With Cervical Spinal Cord Injury | Neuralink has received Health Canada approval to begin recruiting for this clinical trial in Canada. |
| SR014 | University Health Network | UHN performs first Neuralink implant surgeries outside the U.S. | These procedures mark the first Neuralink surgeries performed outside the United States. |
| SR015 | Economy Middle East | Neuralink launches UAE-PRIME, first Middle East clinical trial in Abu Dhabi | In partnership with Abu Dhabi’s Department of Health, the UAE-PRIME clinical trial aims to explore how individuals with motor and speech impairment can use thought to control devices and communicate. |
| SR016 | ClinicalTrials.gov | Study Details | NCT06992596 | UAE-PRIME: A Feasibility Study of a Precise Robotically Implanted Brain-Computer Interface for the Control of External Devices | This study has 1 location ... Cleveland Clinic Abu Dhabi (CCAD). |
| SR017 | Neuralink | PRIME Study Progress Update — User Experience | In the weeks following the surgery, a number of threads retracted from the brain, resulting in a net decrease in the number of effective electrodes. |
| SR018 | Neuralink | Two Years of Telepathy | With a growing crew of 21 Neuralnauts enrolled in trials worldwide, we are making exciting progress. |
| SR019 | Neuralink | Neuralink raises $650 million Series E | Five individuals with severe paralysis are now using Neuralink to control digital and physical devices with their thoughts. |
| SR020 | MedTech Dive | Neuralink raises $650M to help expand patient access | Neuralink also announced that its brain implant has entered clinical trials in three countries, and that five people with severe paralysis are now using the implant. |
| SR021 | Petrie-Flom Center at Harvard Law School | Mind over Machine: Navigating the Legal and Ethical Frontier of Neurotech | Neural data, capturing thoughts, emotions, and predispositions, is perhaps the most intimate form of personal information. |
| SR022 | Voices in Bioethics / Columbia University | The Ethical Significance of Brain-Computer Interfaces as Enablers of Communication | Device failure or abandonment inflicts harm that transcends mere technical malfunction. |
| SR023 | arXiv | Regulating Next-Generation Implantable Brain-Computer Interfaces: Recommendations for Ethical Development and Implementation | Existing regulatory frameworks designed for implantable medical devices are inadequate to address the unique ethical, legal, and social risks associated with next-generation networked brain-computer interfaces. |
| SR024 | KU Leuven Faculty of Law | Neural Privacy and Democratic Self-Rule: Governance Challenges of Invasive Brain-Computer Interfaces | BCIs mark a novel governance challenge where private firms gain unprecedented access to neural data. |
| SR025 | Frontiers in Human Dynamics | Neuralink’s brain-computer interfaces: medical innovations and ethical challenges | Neuralink’s advancements in brain-computer interface technology have positioned the company as a leader in this emerging field. |
| SR026 | U.S. Government Accountability Office | On the Horizon: Three Science and Technology Trends That Could Affect Society | Neural implants for human augmentation could enable direct brain-to-brain communication ... but could also compromise user privacy and security. |
| SR027 | ClinicalTrials.gov | Study Details | NCT06429735 | Precise Robotically IMplanted Brain-Computer InterfacE | The N1 Implant is a skull-mounted, wireless, rechargeable implant connected to electrode threads that are implanted in the brain by the R1 Robot. |
| SR028 | Neuralink | Privacy Policy | Our privacy practices in connection with clinical trials are governed by applicable clinical trial protocols and, where applicable, HIPAA authorization forms. |
| SR029 | Neuralink | Clinical Trials | Join our Patient Registry for upcoming trials. |
| SR030 | U.S. Government Accountability Office | Science & Tech Spotlight: Brain-Computer Interfaces | The technology remains largely experimental, and it raises questions about security, ethics, and equity. |
| SV001 | CNBC / Reuters | Musk's Neuralink valued at about $5 billion despite long road to market: Reuters | Such so-called secondary trades are an imperfect gauge of a company's value; their volume is thin and they lack the wider market consensus of a fundraising round or IPO. |
| SV002 | Semafor | Exclusive: Elon Musk’s Neuralink raises fresh cash at $9B valuation | Neuralink raised $600 million in a deal that values the brain-mapping company at $9 billion before the new cash. |
| SV003 | CNBC | Elon Musk's Neuralink raises $650 million in fresh capital | Elon Musk's Neuralink raises $650 million in fresh capital. |
| SV004 | TechCrunch | Elon Musk's Neuralink closes a $650M Series E | Semafor reported last week that this latest deal values Neuralink at around $9 billion pre-money. |
| SV005 | Fierce Biotech | Neuralink secures $650M series E funding to expand patient access to brain chip technology | Semafor reported last week that the latest deal values Neuralink at $9 billion pre-money. |
| SV006 | Neuralink | Neuralink raises $650 million Series E | The funds will help us bring our technology to more people— restoring independence for those with unmet medical needs and pushing the boundaries of what’s possible with brain interfaces. |
| SV007 | Sacra | Neuralink valuation, funding & news | |
| SV008 | PM Insights | Neuralink Valuation | PM Insights | |
| SV009 | Premier Alternatives | Neuralink - Private Company Valuation & Stock Data | |
| SV010 | Premier Alternatives | Neuralink Valuation: $9.7B (2026) | Neuralink is currently valued at $14.9B as of June 2, 2025. |
| SV011 | Acquinox Capital | Neuralink: Investor Insights into the BCI Opportunity | As of early 2026, secondary market activity suggests implied valuations approaching $40 billion, roughly four times the last formal round, though this figure remains unconfirmed and should be treated as indicative. |
| SV012 | Inside BCI | Neuralink hires its first federal lobbyists to open the brain-computer interface coverage account | The “coverage” framing is the technical phrasing that maps directly to Medicare, Medicaid, and commercial-payer reimbursement determinations. |
| SV013 | U.S. Government Accountability Office | Brain-Computer Interfaces: Applications, Challenges, and Policy Options | Some clinical trial participants have had a BCI removed because there were no funds or medical support provided after the trial. |
| SV014 | Centers for Medicare & Medicaid Services | New Medical Services and New Technologies | Technologies under NTAP review that have not received FDA marketing authorization by the deadline will no longer be eligible for NTAP or discussed in the final rule. |
| SV015 | STAT | What does Neuralink want — to help people with paralysis, or prepare for a war with AI? | This conflicting messaging from a company perceived as the emerging field’s leader could hinder the ability of startups developing brain-computer interfaces to gain approval for them as medical devices and be paid for by health insurers. |
| SV016 | Fast Company | FDA urged Musk’s Neuralink animal lab to fix its objectionable conditions | FDA urged Musk’s Neuralink animal lab to fix its objectionable conditions. |
| SV017 | Yahoo Finance / Bloomberg | Former Neuralink Staffer Sues After Scratches From Herpes Monkey | Former Neuralink Staffer Sues After Scratches From Herpes Monkey. |
| SV018 | HR Dive | Ex-Neuralink staffer claims she was scratched by infected monkeys, fired over her pregnancy | Ex-Neuralink staffer claims she was scratched by infected monkeys, fired over her pregnancy. |
| SV019 | Physicians Committee for Responsible Medicine | SEC Reopens Investigation Into Elon Musk’s Neuralink Likely Launched by Medical Ethics Group’s Complaint About Monkey Deaths | SEC Reopens Investigation Into Elon Musk’s Neuralink Likely Launched by Medical Ethics Group’s Complaint About Monkey Deaths. |
| SV020 | Business Wire / Synchron | Synchron Raises $200 Million Series D to Advance Brain-Computer Interface Technology | The financing brings Synchron’s total funding to $345 million. |
| SV021 | MassDevice | Synchron raises $200M Series D to support BCI commercialization, next-gen tech | It has been implanted in 10 patients with paralysis to date across trials in the U.S. and Australia. |
| SV022 | Yahoo Finance / Globe Newswire | Precision Neuroscience Raises $102 Million to Advance AI-Powered Brain Implant | With $155 million in total capital raised, Precision is now one of the best-funded companies in the BCI industry. |
| SV023 | National Law Review / Precision Neuroscience | Precision Neuroscience Expands Clinical Research in Brain–Computer Interface Through Collaboration with Beth Israel Deaconess Medical Center | |
| SV024 | Securities and Exchange Commission | NeuroPace, Inc. Annual Report on Form 10-K for fiscal year ended December 31, 2025 | The Company’s trailing 12-month net RNS System revenue was $81.7 million as of December 31, 2025. |
| SV025 | Stock Analysis | NeuroPace (NPCE) Market Cap & Net Worth | NeuroPace has a market cap or net worth of $533.23 million as of June 12, 2026. |
| SV026 | Inspire Medical Systems | Inspire Medical Systems, Inc. Announces Fourth Quarter and Full Year 2025 Financial Results | Revenue increased 14% to $912.0 million. |
| SV027 | Macrotrends | Inspire Medical Systems Market Cap 2017-2025 | INSP | Inspire Medical Systems market cap as of January 20, 2026 is $2.74B. |
| SV028 | Medtronic | Medtronic reports fourth quarter and full year fiscal 2026 results; delivers highest annual revenue growth in 10 years | FY26 revenue of $36.4 billion, adjusted revenue of $36.3 billion, increased 8.4% as reported and 5.8% organic. |
| SV029 | CompaniesMarketCap | Medtronic (MDT) - Market capitalization | As of June 2026 Medtronic has a market cap of $102.96 Billion USD. |
| SV030 | Abbott | Abbott Reports Fourth-Quarter and Full-Year 2025 Results; Issues 2026 Financial Outlook | Full-year 2025 sales of $44.3 billion increased 5.7 percent on a reported basis. |
| SV031 | CompaniesMarketCap | Abbott Laboratories (ABT) - Market capitalization | As of June 2026 Abbott Laboratories has a market cap of $153.59 Billion USD. |
| SV032 | Multiples.vc | Multiples Medical Devices Index | Data as of June 12, 2026. |