OpenAI is the leading multi-surface frontier AI platform in this source set, but current underwriting is constrained more by valuation, governance, compute economics, and investor-rights opacity than by product or demand.
OpenAI is the leading multi-surface frontier AI platform in this source set, but current underwriting is constrained more by valuation, governance, compute economics, and investor-rights opacity than by product or demand.
The refresh adds usable but imperfect market anchors: CNBC's March 2025 OpenAI financing at about 30x June 2025 reported ARR, SoftBank's 2026 follow-on at roughly 31.7x the later annualized revenue run-rate, and Anthropic's September 2025 Series F at about 36.6x disclosed run-rate revenue. CoreWeave is a public AI-infrastructure comp for compute economics rather than a model-platform multiple. Using OpenAI's low-confidence $24 billion annualized revenue run-rate from the company-reported March 2026 monthly revenue claim, the disclosed $852 billion post-money valuation implies about 35.5x revenue, broadly consistent with frontier-AI private comps but still demanding given missing audited margins, retention, burn, and customer concentration.
OpenAI looks operationally strong but institutionally complex. The biggest risks are not lack of demand; they are valuation paid against opaque economics, mission-controlled governance with limited investor leverage, leadership concentration, financing-term complexity, Microsoft revenue-share obligations, and the capital intensity of frontier AI infrastructure.
Audited segment revenue, gross margin, retention, and customer concentration across consumer, enterprise, API, and government remain undisclosed.
Market: OpenAI is pursuing a very large and expanding market across consumer AI, developer tooling, enterprise workflow automation, education, and government. The combination of agentic product expansion, enterprise controls, partner distribution, and public-sector procurement creates a strong why-now. The main caveat is that many top-line scale claims remain self-reported and public unit-economics disclosure is still insufficient for high-confidence underwriting.
Product: OpenAI has a scaled, multi-surface product platform with credible enterprise hardening, workflow breadth, and real evidence of production use beyond experimentation. The key caveat is that breadth should not be confused with an unassailable moat: partner-cloud dependence, policy limits in high-stakes domains, and rollout complexity all matter to product durability.
Traction: OpenAI shows unusually broad demand across consumer, enterprise, developer, and government channels. The strongest evidence is a mix of company-reported scale claims such as more than 5 million business users, more than 900 million weekly users, more than 50 million subscribers, and about $2 billion of monthly revenue, plus independently corroborated customer deployment evidence and procurement or distribution expansion through GSA, AWS, and Azure-linked channels. Confidence is capped because the largest usage and revenue figures are self-reported and segment-level monetization remains opaque.
Economics: OpenAI appears to have real and unusually large monetization across subscriptions, enterprise seats, and API usage, now supported by an independently reported $10B ARR milestone and paying-business-user growth. However, public financial transparency is still too limited for high-confidence underwriting: gross margin, current burn, retention, capital efficiency, customer concentration, and segment mix remain opaque while frontier-model and AI-infrastructure commitments appear structurally capital intensive.
OpenAI remains strategically important and continues to grow, but governance complexity, cloud dependence, and incomplete economic disclosure keep valuation multiples from expanding meaningfully beyond an already elevated base.
OpenAI converts its broad product footprint, strong brand pull, and enterprise and government distribution into a deeper workflow platform that compounds across consumer, developer, enterprise, and public-sector budgets.
The company sustains impressive usage but disappoints on monetization quality or cost structure, while governance constraints and competition reduce flexibility just as the current valuation leaves little margin for error.
OpenAI clears the product, market, and strategic-importance bars, and the targeted refresh strengthens the traction file with $10B ARR, 3M paying business users, and additional financing/comparable evidence. It still does not clear the underwriting bar for a large new-money position at the disclosed valuation because gross margin, retention, customer concentration, compute economics, and investor-rights details remain insufficiently disclosed.
Category-defining product breadth across consumer, enterprise, developer, and government surfaces.
Customer, partner, and procurement evidence indicates adoption is moving beyond pilots.
Independent reporting supports a $10B ARR milestone and 3M paying business users in 2025.
Enterprise trust packaging appears more mature than most AI application startups.
Core revenue-quality metrics remain undisclosed despite extraordinary reported scale.
Competitive differentiation is real but not obviously unassailable across every workload.
Leadership continuity and organizational complexity remain active monitoring items.
Mission-first governance gives outside investors limited conventional control leverage.
The valuation already prices in exceptional execution while public unit-economics evidence is thin.
Capital intensity, Microsoft revenue-share obligations, and Stargate/CoreWeave-style commitments could keep equity value capture below headline demand growth.
| Thesis | Why it matters | Status | Evidence |
|---|---|---|---|
| Enterprise, API, and government monetization can compound faster than consumer or promotional usage while improving blended gross margin. | This determines whether the $852B valuation is anchored to durable, high-quality revenue rather than headline usage. | Partially Proven | [ C013, C018, C022, C023, C038, C039, C054, C055 ] |
| Compute, safety, and partner-cloud costs decline or scale efficiently enough to create eventual operating leverage. | Frontier-AI revenue can be strategically valuable while still producing weak equity returns if infrastructure economics consume the upside. | Unproven | [ C031, C033, C039, C044, C057, C060, C062, C064 ] |
| Foundation control and partner influence remain compatible with investor transparency and downside protection. | Governance determines whether outside capital can monitor and protect value through strategic trade-offs. | Partially Proven | [ C002, C003, C004, C005, C036, C051, C053, C058 ] |
The reported revenue scale is substantially paid, retained, and expansion-led rather than mostly low-margin or promotional usage.
OpenAI can maintain frontier product leadership while reducing inference cost per unit of value delivered.
Multi-cloud and public-sector distribution broaden access without weakening OpenAI's economics or control over customer relationships.
Audited or board-level materials show materially weaker gross margin, retention, or customer concentration than implied by the disclosed valuation.
New investors cannot obtain sufficient information rights, pro rata rights, or clarity on economic alignment under Foundation control.
Compute commitments, Microsoft revenue-share obligations, or partner financing mechanics make cash needs materially larger than public materials imply.
High: OpenAI has unusually broad product and channel coverage across consumer, enterprise, APIs, AWS, Azure-linked paths, and government procurement. [ C012, C031, C050 ]
High: Customer and procurement evidence suggests deployment is moving beyond pilots in finance, internet, operations, and U.S. government. [ C018, C023, C029, C054 ]
Medium: Enterprise trust packaging is materially more mature than lightweight AI tools, with no-training defaults, compliance artifacts, and admin controls. [ C010, C015, C042 ]
Medium: Hiring intensity suggests the company is still investing heavily against research, infrastructure, safety, and monetization demand. [ C011 ]
High: The most important revenue, user, and valuation figures are self-reported, while audited economics remain undisclosed. [ C038, C039 ]
High: Governance is unusually mission-controlled and gives outside investors limited conventional leverage. [ C002, C003, C005 ]
High: Key-person dependence and 2024 turnover indicate nontrivial execution fragility. [ C045, C046 ]
Medium: Capital intensity and cloud-partner dependence can keep value capture below headline usage growth. [ C031, C033, C044, C057 ]
High: SoftBank's preferred-share follow-on and bridge financing improve valuation evidence but underscore the need to diligence preferences, conversion mechanics, and investor rights. [ C053, C058, C059 ]
OpenAI shows unusually broad demand across consumer, enterprise, developer, and government channels. The strongest evidence is a mix of company-reported scale claims such as more than 5 million business users, more than 900 million weekly users, more than 50 million subscribers, and about $2 billion of monthly revenue, plus independently corroborated customer deployment evidence and procurement or distribution expansion through GSA, AWS, and Azure-linked channels. Confidence is capped because the largest usage and revenue figures are self-reported and segment-level monetization remains opaque.
ChatGPT offers free and paid consumer tiers, ChatGPT Business is publicly listed at SGD 25 per user per month billed annually, Enterprise is custom priced, API pricing is usage based, and U.S. federal agencies have a $1 first-year OneGov entry offer for ChatGPT Enterprise.
OpenAI packages separate consumer, SMB, enterprise, education, government, and developer surfaces. Enterprise value is bundled with shared workspaces, connectors, admin controls, retention settings, and deployment support rather than raw model access alone.
Morgan Stanley reported daily usage across nearly all advisor teams and over 98% adoption in wealth management.
CyberAgent reported a 93% monthly active user rate across nearly all departments.
Enterprise packaging emphasizes connectors, shared workspaces, apps, and admin controls that can support seat expansion and deeper workflow embedment.
| Type | Signal | Strength | Independent | Evidence |
|---|---|---|---|---|
| Revenue | OpenAI's March 2026 financing post claimed about $2 billion in monthly revenue and enterprise revenue above 40% of total. | High | No | [ C038 ] |
| Users | OpenAI's March 2026 financing post claimed more than 900 million weekly users and over 50 million subscribers. | High | No | [ C038 ] |
| Customers | OpenAI says ChatGPT Enterprise serves more than 5 million business users across industries. | High | No | [ C022 ] |
| Usage | Morgan Stanley reported over 98% advisor-team adoption and CyberAgent reported a 93% monthly active user rate across nearly all departments, suggesting deployments are moving beyond pilot usage. | Medium | Yes | [ C023 ] |
| Partnerships | OpenAI distribution now spans Azure-linked channels, AWS Bedrock-managed delivery, and GSA OneGov procurement, widening enterprise and public-sector reach. | High | Yes | [ C018, C031, C050 ] |
| Usage | ChatGPT Gov materials claim more than 90,000 users across more than 3,500 U.S. agencies and over 18 million messages since 2024. | Medium | No | [ C016 ] |
| Hiring | OpenAI's careers page listed 669 open roles across research, infrastructure, safety, government, and monetization functions, implying continued investment ahead of demand. | Medium | No | [ C011 ] |
| Customer | Industry | Scope | Outcome | Verified | Evidence |
|---|---|---|---|---|---|
| Morgan Stanley | financial services | Wealth-management advisor teams using internal AI assistant workflows | Nearly all advisor teams use the assistant daily, with over 98% adoption in wealth management. | No | [ C023, C024 ] |
| CyberAgent | internet and digital media | Organization-wide ChatGPT Enterprise and Codex usage across nearly all departments | Monthly active user rate reached 93% after training and governance rollout. | No | [ C023, C025 ] |
| Choco | food distribution software | AI-powered ordering and voice workflows embedded in distributor operations | Processes more than 8.8 million orders annually, cuts manual order entry by up to 50%, and doubles sales productivity without added headcount. | Yes | [ C023, C028, C029 ] |
| Gradient Labs | financial-services support software | AI account-manager and support workflows for banks and financial institutions | Customer-side materials cite 80+ CSAT, up to 98% CSAT in optimal implementations, and 40 to 60% day-one resolution rates. | Yes | [ C023, C026, C027 ] |
The market is shifting from experimentation to managed deployment because capability, governance, and buying paths have matured together. OpenAI is now selling not only model access but also secure workspaces, sector packages, partner distribution, and procurement-ready offerings.
Technology: Agentic AI is becoming useful for multi-step work, not just Q and A, enabling coding, research, support, and operational execution inside managed products.
Distribution: OpenAI now has more mature distribution and deployment rails through direct SaaS, APIs, Azure or Azure Government, AWS, and GSA procurement.
Regulatory: Buyer trust has improved because privacy, retention, compliance, and admin controls are now productized rather than promised abstractly.
Behavioral: Business adoption no longer starts from zero familiarity because OpenAI claims millions of business users and customer cases show broad internal usage once governance and training are in place.
Economic: Early customer evidence now includes measurable time, quality, service, and productivity gains in finance, internet, and operations-heavy workflows.
OpenAI is pursuing a very large and expanding market across consumer AI, developer tooling, enterprise workflow automation, education, and government. The combination of agentic product expansion, enterprise controls, partner distribution, and public-sector procurement creates a strong why-now. The main caveat is that many top-line scale claims remain self-reported and public unit-economics disclosure is still insufficient for high-confidence underwriting.
| Scope | Value | USD M | Approach | Methodology | Confidence |
|---|---|---|---|---|---|
| TAM | Not disclosed | Not disclosed | Unknown | No defensible single TAM figure is used in this standard-depth pass because OpenAI spans consumer subscriptions, enterprise software, developer infrastructure, and public-sector procurement. Public disclosures do not support a clean, non-double-counted roll-up. | Low |
| SAM | Not disclosed | Not disclosed | Bottom Up | The practical near-term SAM is the subset of knowledge-work, software-development, customer-support, education, and public-sector workflows that can adopt secure ChatGPT, Codex, or API deployments under current compliance, budget, and policy constraints. | Medium |
| SOM | Not disclosed | Not disclosed | Bottom Up | Current capture is better represented by disclosed usage and deployment signals than by revenue segmentation. Public evidence shows business users, customer deployments, and government activity, but not enough audited economics to quantify a reliable SOM. | Medium |
| Segment | Buyer | Budget owner | Urgency | WTP evidence | Evidence |
|---|---|---|---|---|---|
| Consumers and prosumers | Individual user | Individual discretionary budget | Medium | ChatGPT public pricing shows paid self-serve tiers from Go through Pro. | [ C007, C008, C012, C041 ] |
| Developers and AI-native product teams | CTO, VP Engineering, or product leader | Engineering or product budget | High | OpenAI API pricing is explicit, and OpenAI plus AWS now support Codex and model usage in AWS environments while Microsoft remains a major channel. | [ C007, C009, C030, C031, C033, C041 ] |
| SMBs and mid-market knowledge-work teams | Founder, COO, IT lead, or department head | Operations, IT, or G&A budget | Medium | ChatGPT Business is self-serve and publicly priced at SGD 25 per user per month billed annually. | [ C008, C012, C013, C041 ] |
| Large enterprises | CIO, CDO, head of AI, or enterprise function leader | Central IT or digital-transformation budget with business-unit sponsors | High | OpenAI claims more than 5 million business users, and enterprise case studies show high adoption plus deployment support expectations. | [ C010, C013, C022, C023, C025, C042 ] |
| Financial-services institutions | CIO, chief data officer, head of AI, or compliance-aligned business leader | Technology, operations, or line-of-business transformation budget | High | OpenAI markets a finance-specific solution, Morgan Stanley reports broad adoption, and Gradient Labs positions finance AI support as a production category. | [ C014, C023, C024, C026, C027, C042 ] |
| Educational institutions | CIO, provost, or campus IT administrator | Central IT or academic technology budget | Medium | ChatGPT Edu is marketed as a managed campus offering with no-training defaults, SAML, SCIM, and retention controls. | [ C015, C041, C042 ] |
| Government and public-sector agencies | Agency CIO, procurement lead, CDAO, or mission-program owner | Agency IT, procurement, or mission-program budget | Medium | OpenAI markets ChatGPT Gov and FedRAMP paths, and GSA independently announced a $1 ChatGPT Enterprise offer for participating U.S. federal agencies. | [ C016, C017, C018, C019, C041, C050 ] |
| Operations-heavy enterprises embedding AI into frontline workflows | COO, VP Operations, head of support, or product owner | Operations or business-process transformation budget | High | Choco and Gradient Labs cite production automation, reduced manual work, and customer-service outcomes in operations-heavy settings. | [ C023, C026, C027, C028, C029 ] |
OpenAI has a scaled, multi-surface product platform with credible enterprise hardening, workflow breadth, and real evidence of production use beyond experimentation. The key caveat is that breadth should not be confused with an unassailable moat: partner-cloud dependence, policy limits in high-stakes domains, and rollout complexity all matter to product durability.
The public architecture reads as a layered platform rather than a single model endpoint: direct ChatGPT products sit above developer APIs and coding or agent surfaces, while enterprise value depends on connecting customer data and tools under admin, retention, and compliance controls. The stack is becoming more multi-cloud in distribution, but it still carries meaningful partner-cloud dependence because Microsoft remains the primary cloud partner even as AWS-based delivery expands.
OpenAI says business and API customer data is not used for training by default and is protected with encryption, retention controls, admin features, and enterprise identity support. Public materials also cite DPA or BAA support, audit visibility, explicit usage-policy restrictions for sensitive deployments, and board-level safety and security oversight; at the same time, trust disclosures show moderation and legal-process obligations are already material at scale.
| Item | Timeframe | Source | Evidence |
|---|---|---|---|
| OpenAI models, Codex, and managed agents are being distributed inside AWS environments | Shipped | Announcement | [ C031 ] |
| OpenAI can now serve products across any cloud provider while Azure remains the primary cloud partner and first-ship destination | Shipped | Announcement | [ C033 ] |
| Government packaging continues to harden around ChatGPT Gov, FedRAMP 20x Moderate positioning, and AWS GovCloud deployment paths | Near | Announcement | [ C017 ] |
| Hiring remains heavy across datacenter design, safety, investigations, forward-deployed engineering, government, monetization, and finance | Near | Careers | [ C011 ] |
Official materials fetched for this run do not disclose a primary-source total headcount or functional mix. The official careers search page listed 669 open roles on 2026-05-01 across research, applied engineering, safety, security, model deployment, government, monetization, hardware, and datacenter infrastructure roles in multiple geographies, indicating continued rapid buildout.
OpenAI presents a high-caliber founder and board mix and unusually broad active hiring across research, product, deployment, safety, infrastructure, and government functions. However, investor-grade visibility into the operating org chart, total headcount, and succession depth remains incomplete, and key-person dependence on Sam Altman and a narrow senior bench remains a material consideration.
OpenAI operates through a Foundation-plus-PBC structure in which the OpenAI Foundation governs OpenAI Group PBC, appoints all Group directors, and oversees safety through an independent board committee.
Economic upside exists for investors, but governance is explicitly mission-first and board control does not follow ordinary venture norms; Microsoft also already holds a large stake, and disclosed SoftBank preferred-share mechanics require document-level review.
Co-chair at founding; CEO
CurrentCo-chaired OpenAI at founding and remains the organization's CEO and a board member under the current structure.
Co-founder; President
CurrentFormer CTO of Stripe and a core OpenAI founder-operator who continues to lead the organization alongside Sam Altman.
Co-founder; Head of AI Resilience at OpenAI Foundation
CurrentFounding research leader who, as of March 2026, is leading AI Resilience work within the OpenAI Foundation.
Co-founder; founding research director; former chief scientist
FormerServed as OpenAI's founding research director and later chief scientist before his 2024 departure.
Co-founder; former alignment science leader
FormerLed reinforcement learning and alignment-related work and departed for Anthropic in 2024.
Chief Scientist
Currentresearch
Head of Life Sciences and Curing Diseases
CurrentCoefficient Giving
Head of AI for Civil Society and Philanthropy
CurrentOpenAI
| Name | Role | Current | Domain fit | Background |
|---|---|---|---|---|
| Sam Altman | Co-chair at founding; CEO | Yes | High | Co-chaired OpenAI at founding and remains the organization's CEO and a board member under the current structure. |
| Greg Brockman | Co-founder; President | Yes | High | Former CTO of Stripe and a core OpenAI founder-operator who continues to lead the organization alongside Sam Altman. |
| Wojciech Zaremba | Co-founder; Head of AI Resilience at OpenAI Foundation | Yes | High | Founding research leader who, as of March 2026, is leading AI Resilience work within the OpenAI Foundation. |
| Ilya Sutskever | Co-founder; founding research director; former chief scientist | No | High | Served as OpenAI's founding research director and later chief scientist before his 2024 departure. |
| John Schulman | Co-founder; former alignment science leader | No | High | Led reinforcement learning and alignment-related work and departed for Anthropic in 2024. |
| Name | Role | Function | From | Significance |
|---|---|---|---|---|
| Jakub Pachocki | Chief Scientist | Research | Not disclosed | High |
| Jacob Trefethen | Head of Life Sciences and Curing Diseases | Research | Coefficient Giving | Medium |
| Anna Makanju | Head of AI for Civil Society and Philanthropy | Other | OpenAI | Medium |
| Robert Kaiden | Chief Financial Officer, OpenAI Foundation | Finance | Not disclosed | Medium |
| Jeff Arnold | Director of Operations, OpenAI Foundation | Operations | Not disclosed | Medium |
Multi-surface AI platform monetizing consumer subscriptions, business and enterprise seats, usage-based APIs, and direct or partner-led sector deployments
Blended gross margin is likely below elite SaaS levels because high-value software revenue is offset by inference-compute intensity, Microsoft revenue-share economics, partner-cloud obligations, safety operations, and potential datacenter or infrastructure costs.
OpenAI appears to have real and unusually large monetization across subscriptions, enterprise seats, and API usage, now supported by an independently reported $10B ARR milestone and paying-business-user growth. However, public financial transparency is still too limited for high-confidence underwriting: gross margin, current burn, retention, capital efficiency, customer concentration, and segment mix remain opaque while frontier-model and AI-infrastructure commitments appear structurally capital intensive.
Public evidence does not disclose ACV or ARPA, cohort retention, CAC, contribution margin, or LTV across consumer, enterprise, API, or government surfaces.
Burn multiple, magic number, and Rule of 40 cannot be computed from the current source set because audited revenue growth, free cash flow, sales efficiency, and current burn are not publicly disclosed. CapitalConsumedToArrRatio is also left null because public sources do not provide a clean cumulative company-wide capital base or consistent ARR definition across periods.
Public evidence now includes an independently reported $10B ARR milestone with a disclosed revenue definition across consumer, business, and API surfaces, plus paying business-user growth. However, the file still lacks audited segment revenue, gross margin, retention, customer concentration, and paid-cohort data. Treat revenue quality as promising but not underwritable without company materials.
| Stream | Quality signal | Notes | Evidence |
|---|---|---|---|
| consumer subscriptions | Medium | Public pricing and the reported ARR definition support monetization, but consumer churn, plan mix, and promotional usage are not disclosed. | [ C007, C008, C038, C039, C055 ] |
| business and enterprise seats | High Potential | Enterprise pages, customer cases, and CNBC's report of 3M paying business users suggest workflow adoption, security controls, and seat-based expansion, but NRR and contract quality remain undisclosed. | [ C010, C013, C022, C023, C042, C054, C055 ] |
| usage-based API | Mixed | Usage-based API pricing is included in the reported ARR definition and can scale with developer demand, but it may carry volatile usage, model substitution risk, and compute-cost sensitivity. | [ C007, C009, C030, C031, C039, C055 ] |
| government and regulated-sector deployments | Emerging | Government procurement and regulated-industry positioning create strategic channels, but public evidence is stronger on access and usage than recurring revenue contribution. | [ C014, C017, C018, C019, C034 ] |
Request enterprise and API cohort retention, expansion ARR, churned-customer analysis, and usage-to-paid conversion by segment.
[ C022, C023, C039 ]Request top-customer revenue concentration, Microsoft/Azure-related revenue exposure, and major API or enterprise customer dependency.
[ C004, C030, C031, C033 ]Request contract duration, prepaid commitments, renewal mechanics, usage-commit structures, and public-sector procurement terms.
[ C013, C018, C019, C034 ]| Scenario | Revenue range | Margin direction | Capital need | Assumptions | Evidence |
|---|---|---|---|---|---|
| Downside | $32 billion to $39 billion | Deteriorating | Potential need for additional external capital, structured compute financing, or tighter spending controls if growth slows while infrastructure commitments remain elevated. | Starts from a low-confidence $24 billion annualized revenue run-rate derived from the company-reported March 2026 $2 billion monthly revenue figure. · Cross-checks against CNBC's June 2025 $10 billion ARR report but assumes growth quality weakens under compute and governance pressure. · Assumes roughly 10% to 18% annualized topline growth over three years because competition, regulation, and monetization quality are less favorable than headline adoption suggests. · Assumes compute, safety, and deployment costs remain heavy relative to revenue growth. | [ C038, C039, C044, C046, C048, C055, C057, C060, C062 ] |
| Base | $41 billion to $54 billion | Improving | The March 2026 financing could support the base case, but ongoing infrastructure and partner-cloud commitments would still keep capital needs material. | Starts from the same low-confidence $24 billion annualized revenue run-rate anchor. · Assumes 2025 reported ARR and paying-business-user growth translate into durable enterprise/API expansion rather than one-off adoption spikes. · Assumes roughly 20% to 31% annualized topline growth as enterprise, developer, and public-sector adoption keep expanding. · Assumes some mix shift toward enterprise and higher-value workflows improves monetization faster than operating costs. | [ C018, C022, C030, C038, C048, C050, C054, C055 ] |
| Upside | $59 billion to $79 billion | Improving | Even in the upside case, OpenAI would likely continue absorbing very large compute and deployment spend, though stronger revenue could partially self-fund expansion. | Starts from the same low-confidence $24 billion annualized revenue run-rate anchor. · Assumes roughly 35% to 49% annualized growth if OpenAI sustains category leadership across consumer, enterprise, developer, and government channels. · Assumes workflow depth and enterprise mix improve enough to offset some compute intensity and partner economics. | [ C018, C022, C031, C038, C041, C048, C050, C054, C055, C057 ] |
The refresh adds usable but imperfect market anchors: CNBC's March 2025 OpenAI financing at about 30x June 2025 reported ARR, SoftBank's 2026 follow-on at roughly 31.7x the later annualized revenue run-rate, and Anthropic's September 2025 Series F at about 36.6x disclosed run-rate revenue. CoreWeave is a public AI-infrastructure comp for compute economics rather than a model-platform multiple. Using OpenAI's low-confidence $24 billion annualized revenue run-rate from the company-reported March 2026 monthly revenue claim, the disclosed $852 billion post-money valuation implies about 35.5x revenue, broadly consistent with frontier-AI private comps but still demanding given missing audited margins, retention, burn, and customer concentration.
The valuation bridge still supports a research-more stance: the price is now better triangulated by private frontier-AI comps, but it requires evidence that revenue quality, compute economics, and investor rights are far stronger than currently visible.
| Bridge step | Value | Interpretation | Evidence |
|---|---|---|---|
| 2025 independently reported ARR anchor | $10B ARR and $300B post-money | CNBC's 2025 reporting implies roughly 30x revenue and establishes a less promotional pre-2026 valuation baseline. | [ C052, C055, C056 ] |
| Company-reported March 2026 revenue run-rate | $24B annualized | Uses $2B monthly revenue × 12; low confidence because it is self-reported and not segmented. | [ C038 ] |
| SoftBank preferred-share follow-on anchor | $730B pre-money / approximately $760B post-money | Investor-disclosed 2026 pricing implies a valuation level close to OpenAI's later $852B company-reported post-money anchor. | [ C058, C059 ] |
| Private frontier-AI comp | Anthropic at about 36.6x disclosed run-rate revenue | Anthropic supports that frontier-AI private rounds can clear at very high revenue multiples, though company-disclosed figures and compute obligations limit comparability. | [ C063, C064 ] |
| Observed post-money valuation | $852B | Treats the March 2026 financing as the clearest current market-clearing anchor. | [ C038 ] |
| Implied revenue multiple | 35.5x | Consistent with private frontier-AI comp multiples but still very demanding unless revenue quality, gross margin, and retention prove exceptional. | [ C038, C039, C063 ] |
| Risk-adjusted sensitivity range | $600B to $1.2T | Wide range reflects exceptional strategic position offset by limited public evidence on unit economics, governance, and compute obligations. | [ C039, C044, C048, C057, C060, C062 ] |
| Name | Ticker/Event | Rationale | Revenue multiple | Valuation | Evidence |
|---|---|---|---|---|---|
| Not disclosed | CRWV | Not disclosed | Not disclosed | Not disclosed | [ C061, C062 ] |
| Not disclosed | Not disclosed | Not disclosed | 36.6 | Not disclosed | [ C063 ] |
| Not disclosed | Not disclosed | Not disclosed | 30 | Not disclosed | [ C052, C055, C056 ] |
| Not disclosed | Not disclosed | Not disclosed | 31.7 | Not disclosed | [ C038, C058, C059 ] |
| Not disclosed | Not disclosed | Not disclosed | Not disclosed | Not disclosed | [ C064 ] |
| Scenario | Exit value | Year | Type | MoM | IRR | Assumptions | Evidence |
|---|---|---|---|---|---|---|---|
| Downside | $600B | 2030 | Writedown | 0.704 | -8.4% | Assumes the current round proves too aggressive relative to durable margins and growth, leading to multiple compression even if OpenAI remains strategically important. | [ C038, C039, C044, C046 ] |
| Base | $1,200B | 2030 | Ipo | 1.408 | 8.9% | Assumes OpenAI sustains leadership and grows meaningfully from the current run-rate, but valuation expansion is limited because margins and governance remain contentious for public investors. | [ C022, C031, C038, C039, C048 ] |
| Upside | $2,000B | 2030 | Ipo | 2.347 | 23.8% | Assumes OpenAI compounds leadership across consumer, enterprise, developer, and government channels while proving better monetization quality than public evidence currently shows. | [ C018, C022, C031, C038, C041, C048, C050 ] |
| Valuation | Check size | Ownership | Required exit for 3x | Interpretation |
|---|---|---|---|---|
| $600B | $85M | 0.014% | $1,800B | A materially lower entry valuation creates a plausible path to venture-style upside if OpenAI reaches public-market mega-cap scale. |
| $852B | $85M | 0.01% | $2,556B | At the disclosed anchor, a 3x gross outcome requires roughly $2.6T of exit value before dilution. |
| $1,200B | $85M | 0.007% | $3,600B | Upside entry pricing would require extraordinary long-term public-market outcomes to justify new-money risk. |
Ipo / Medium: Earliest year 2,030
Governance complexity and Foundation control may require clearer public-market narrative. · Audited margin, retention, and capex profile would need to support a mega-cap valuation. · Compute commitments, Microsoft revenue-share economics, and investor preferences would need transparent public-market disclosure.
[
C002, C003, C038, C039
]
Strategic Acquisition / Low: Earliest year Not disclosed
Scale, national-security salience, and competition scrutiny make acquisition by a major platform difficult. · Microsoft ownership and partnership terms would complicate any alternative strategic process.
[
C004, C032, C033, C034
]
Secondary Liquidity / High: Earliest year 2,027
Pricing may depend heavily on continued access demand despite limited financial disclosure. · Information rights and transfer restrictions could constrain buyer universe.
[
C038, C039
]
IPO is the cleanest long-term liquidity path, but public-market readiness depends on converting strategic importance into audited financial quality and a governance story public investors can underwrite.
Public valuation and revenue narrative are far easier to observe than underlying unit economics, leaving investors exposed to multiple compression, margin disappointment, or cash-flow drag from compute and revenue-share obligations.
Foundation control and mission-first duties limit ordinary investor influence over board composition, strategy, and downside management.
Key-person dependence remains high because leadership credibility, fundraising, and external trust are concentrated in Sam Altman and a small senior bench amid recent turnover.
Cloud and channel dependence remain material even after AWS expansion because Microsoft is still the primary cloud partner and first-ship destination.
Moderation load, usage-policy limits, and past antitrust scrutiny show regulatory and policy obligations can shape product scope and go-to-market.
Enterprise and public-sector adoption depend on procurement, compliance, connectors, and change management, which can slow monetization even when demand is strong.
| ID | Category | Risk | Severity | Likelihood | Mitigation | Evidence |
|---|---|---|---|---|---|---|
| R001 | Financial | Public valuation and revenue narrative are far easier to observe than underlying unit economics, leaving investors exposed to multiple compression, margin disappointment, or cash-flow drag from compute and revenue-share obligations. | Critical | High | Require audited or otherwise credible disclosure of segment revenue, gross margin, burn, retention, and customer concentration before underwriting price. | [ C038, C039, C055, C056, C057, C060, C062 ] |
| R002 | Governance | Foundation control and mission-first duties limit ordinary investor influence over board composition, strategy, and downside management. | High | High | Treat this as a governance-light position and seek strong information rights and pro rata rather than expecting conventional control. | [ C002, C003, C004, C005 ] |
| R003 | Execution | Key-person dependence remains high because leadership credibility, fundraising, and external trust are concentrated in Sam Altman and a small senior bench amid recent turnover. | High | Medium | Test succession depth, delegated operating ownership, and retention plans across research, product, safety, and go-to-market. | [ C045, C046 ] |
| R004 | Platform | Cloud and channel dependence remain material even after AWS expansion because Microsoft is still the primary cloud partner and first-ship destination. | High | Medium | Monitor whether multi-cloud execution materially diversifies revenue, deployment, and bargaining power. | [ C031, C032, C033, C050 ] |
| R005 | Regulatory | Moderation load, usage-policy limits, and past antitrust scrutiny show regulatory and policy obligations can shape product scope and go-to-market. | High | Medium | Prioritize regulated-use-case governance, documented evals, and periodic review of transparency, policy, and competition disclosures. | [ C020, C021, C034, C035, C042, C046 ] |
| R006 | Gtm | Enterprise and public-sector adoption depend on procurement, compliance, connectors, and change management, which can slow monetization even when demand is strong. | High | High | Track paid expansion, procurement conversion, and deployment depth rather than top-of-funnel usage alone. | [ C019, C025, C042, C050 ] |
| R007 | Competition | OpenAI's breadth is strong, but much of its moat appears to come from distribution and packaging rather than an unassailable isolated product advantage. | Medium | High | Look for evidence that connectors, workflow embedment, and multi-channel distribution are producing durable retention or higher share of wallet. | [ C030, C031, C033, C048 ] |
| R008 | Funding | Frontier-model development and deployment remain structurally capital intensive, so even very large revenue can coexist with heavy ongoing capital needs, supplier commitments, and structured financing complexity. | High | High | Obtain visibility into compute commitments, Microsoft economics, infrastructure financing, SPVs, bridge facilities, and scenario-level cash needs. | [ C044, C046, C057, C060, C062, C064 ] |
| R009 | Financing Terms | Financing conditionality, automatically converting preferred shares, convertible notes, and bridge-backed follow-on funding may create preference, conversion, and information-rights issues that are not visible in public sources. | High | Medium | Review definitive financing documents, liquidation preferences, conversion triggers, pro rata rights, information rights, and any restructuring contingencies before investment. | [ C051, C053, C058, C059 ] |
High: Unit economics prove materially weaker than implied by the March 2026 valuation anchor.
Audited or well-sourced disclosures show weaker gross margin, heavier burn, or lower enterprise mix than expected. · Revenue growth decelerates without corresponding improvement in retention or ARPA. · New financings occur on less favorable terms despite continued usage growth. · Infrastructure commitments, revenue-share obligations, or supplier contracts consume more cash than revenue growth offsets.
Medium: Mission-first governance and concentrated control limit corrective action when strategic or financial trade-offs worsen.
Outside investors receive limited incremental disclosure. · Board or committee changes reduce perceived independence. · Major strategic decisions appear driven by partnership or mission constraints rather than shareholder value optimization.
Medium: Leadership turnover or bench weakness slows product, safety, or commercial execution.
Additional senior departures. · Roles remain open across critical safety, infrastructure, or deployment functions. · Public messaging around ownership of research, product, or Foundation priorities becomes more fragmented.
Medium: Competition and partner dependence compress pricing power or distribution leverage.
Key workloads migrate to customer-cloud alternatives. · OpenAI's multi-cloud expansion fails to broaden procurement win rates. · Customer evidence emphasizes cloud-boundary fit over OpenAI-specific workflow value.
Upgrade / High: Audited disclosure of segment revenue, gross margin, burn, retention, and customer concentration showing strong economics at scale.
Upgrade / High: Independent confirmation that enterprise revenue mix, paid conversion, and cohort durability support the March 2026 scale claims.
Upgrade / Medium: Evidence that governance reforms materially improve investor transparency or downside protections without weakening safety oversight.
Upgrade / High: Definitive investor documents showing strong information rights, pro rata rights, and downside protections despite Foundation control and preferred-share mechanics.
Downgrade / High: Credible reporting that large reported usage is materially less monetizable than implied or heavily subsidized by promotions.
Downgrade / High: Further leadership attrition or a major regulatory, safety, or security failure affecting enterprise trust.
Board-approved or auditor-reviewed revenue bridge by segment and geography.
Gross margin and contribution margin by major product surface.
Burn, capex or compute commitments, and cash runway scenario planning.
Top-customer concentration, churn, NRR or GRR, and expansion cohorts.
Current cap-table details, investor rights, and any partnership-related governance constraints.
Definitive terms for SoftBank follow-on preferred shares, convertible notes, conversion triggers, liquidation preferences, and pro rata rights.
Microsoft revenue-share cap, cloud minimums, and economics by cloud/provider route.
Compute commitments by supplier, including Stargate, Oracle, CoreWeave, Microsoft/Azure, and any related SPVs or financing obligations.
Org chart showing ownership across research, product, safety, infrastructure, sales, government, and Foundation activities.
Former or current enterprise buyers who evaluated OpenAI against cloud-native alternatives.
Government procurement specialists familiar with OneGov, FedRAMP, and GovCloud deployment friction.
AI infrastructure experts who can benchmark likely compute intensity and bargaining power under a multi-cloud strategy.
Governance counsel or public-benefit-corporation experts who can assess downside protection under Foundation control.
| Question | Why it matters | Evidence |
|---|---|---|
| What percentage of revenue and gross profit comes from ChatGPT consumer subscriptions, ChatGPT Business or Enterprise, API usage, and government deployments? | Revenue mix is the fastest way to distinguish durable enterprise monetization from lower-quality usage scale. | [ C007, C038, C039 ] |
| What are current NRR, GRR, churn, and expansion cohorts for enterprise and API customers? | Retention quality determines whether OpenAI should be valued like mission-critical software or like a volatile usage platform. | [ C022, C023, C039 ] |
| How do inference cost per dollar of revenue, gross margin, and cloud commitments vary by product surface? | The most important unknown in frontier AI valuation is whether compute intensity falls faster than usage expands. | [ C031, C033, C039, C044, C057, C060, C062 ] |
| What information rights, pro rata rights, and governance protections would new investors receive? | Foundation control and Microsoft ownership make investor-protection terms central to risk-adjusted returns. | [ C002, C003, C004, C005, C051, C053, C058 ] |
| Which workloads are most at risk of substitution by cloud-native models, open-weight models, or customer-built systems? | This tests whether OpenAI's moat is workflow depth and distribution rather than only frontier-model performance. | [ C030, C031, C050 ] |
M001: Provide investor-grade disclosure on segment revenue mix, gross margin, burn, and retention.
↑ Would materially de-risk underwriting by showing whether enterprise and API mix plus unit economics support the current valuation narrative. · ↓ Continued opacity would leave the market relying on self-reported scale claims and keep downside to multiple compression high.
M002: Show that multi-cloud distribution meaningfully expands paid enterprise deployments beyond Azure-linked channels.
↑ Would suggest OpenAI is improving bargaining power and turning AWS availability into real commercial diversification. · ↓ Would imply that cloud optionality is more narrative than economic and that partner concentration remains stubbornly high.
M003: Translate public-sector entry paths such as OneGov, FedRAMP positioning, and GovCloud deployment into durable paid usage rather than mostly low-friction trials.
↑ Would validate government as a meaningful monetization channel rather than mainly a credibility and procurement wedge. · ↓ Would suggest that policy and procurement complexity are diluting the value of the current public-sector narrative.
M004: Demonstrate deeper workflow embedment through connectors, shared workspaces, coding, and agent use cases inside large enterprises.
↑ Would strengthen evidence that OpenAI is becoming workflow infrastructure rather than a replaceable assistant surface. · ↓ Would support the view that adoption is broad but shallow and easier for substitutes to attack.
M005: Reduce key-person and bench risk by keeping senior leadership stable and clarifying ownership across research, safety, infrastructure, and commercial functions.
↑ Would improve confidence that execution can continue without depending excessively on a small founder-level leadership circle. · ↓ Would increase concern that turnover or organizational ambiguity could become a bottleneck during scale-up.
M006: Show that compliance and safety governance can scale without materially slowing product shipping or enterprise trust.
↑ Would support the case that OpenAI can stay aggressive commercially while remaining acceptable to regulated buyers. · ↓ Would indicate that oversight, moderation, or policy burden is becoming a drag on growth or adoption.
M007: Prove that OpenAI's moat is deepening through retention, expansion, and workflow lock-in rather than only brand and distribution breadth.
↑ Would justify a stronger long-duration software multiple if customers become more embedded and harder to displace. · ↓ Would reinforce the concern that the business remains vulnerable to cloud-native, vertical, or internal-build alternatives.
M008: Sustain revenue growth while improving evidence of monetization quality and capital discipline against frontier-AI compute intensity.
↑ Would increase confidence that OpenAI can convert category leadership into durable value capture rather than expensive scale. · ↓ Would make the current valuation harder to justify even if usage remains very high.
M009: Convert multi-channel reach across direct products, AWS, Azure-linked paths, and government procurement into resilient bargaining power and lower partner concentration.
↑ Would show that OpenAI's channel breadth is becoming structural leverage rather than tactical distribution. · ↓ Would leave the company exposed to partner-shaped economics and narrower strategic flexibility than the surface narrative suggests.
| Criterion | Severity | Detection |
|---|---|---|
| Credible evidence shows the March 2026 revenue, user, or subscriber claims materially overstate durable paid demand. | Critical | Audited statements, board-reviewed metrics, or credible investigative reporting |
| Gross margin, burn, or compute commitments imply a structurally poor economic model at the current valuation. | Critical | Financial disclosure or diligence materials |
| Governance changes materially weaken safety oversight or further reduce investor transparency and alignment. | High | Structure, board, or committee changes |
| Further senior-leadership attrition materially degrades execution depth in research, infrastructure, safety, or enterprise sales. | High | Leadership departures and long-open critical roles |
| A major regulatory, security, or trust failure materially impairs enterprise or public-sector deployment. | High | Regulatory actions, customer churn, or incident disclosures |
OpenAI’s current moat looks more like distribution breadth, brand pull, enterprise or government packaging, and growing workflow integration than a clearly unassailable single-product advantage.
If OpenAI turns connectors, admin controls, procurement rails, and eval-driven workflow reliability into durable embedded usage across enterprises and agencies, workflow lock-in could deepen over time.
OpenAI is clearly more than a single chatbot vendor in this evidence set: it combines broad product coverage, regulated-buyer packaging, and expanding channel reach across direct products, partner clouds, and government procurement. Still, the position is only somewhat differentiated rather than impregnable because much of the advantage comes from distribution, workflow fit, and governance packaging instead of a fully proven durable moat.
| Name | Category | Positioning | Strengths | Weaknesses | Evidence |
|---|---|---|---|---|---|
| Alternative frontier-model platforms | foundation-model APIs and enterprise AI surfaces | Compete for developer spend and enterprise AI workflows on model capability, agentic features, and reliability. | Some buyer decisions can swing on narrow workflow reliability and evaluation performance rather than suite breadth · Pure model access can be repackaged through multiple channels, reducing the uniqueness of any one endpoint | This source set shows less evidence of equivalent breadth across consumer ChatGPT, enterprise workspaces, government packaging, and procurement channels · OpenAI already spans direct SaaS, API, coding, and agent surfaces | [ C012, C026, C030, C031, C050 ] |
| Cloud-native managed-agent and model platforms | cloud-distributed enterprise AI platforms | Compete by meeting customers inside existing cloud security, identity, and procurement boundaries rather than pulling them into a separate destination product. | Can reduce deployment friction by aligning with clouds, governance controls, and procurement processes customers already use · Can keep model usage inside existing cloud environments, which matters for certain enterprise and government buyers | Can be intermediary distribution layers rather than full workflow products · OpenAI now reaches customers through both AWS and Azure-linked channels while retaining direct products | [ C019, C031, C033, C050 ] |
| Specialized regulated-workflow AI vendors | vertical AI applications for finance, support, and operations | Compete on domain-specific workflow design, evaluation discipline, and compliance-tuned deployments for narrower use cases. | Can optimize for narrow domain metrics and regulated workflows instead of general-purpose breadth · Customer evidence in finance and operations shows workflow-specific reliability and service metrics matter | Narrower scope can limit cross-budget expansion into general knowledge work, coding, and government · OpenAI’s distribution breadth across direct products, partner clouds, and procurement channels gives it more entry points | [ C023, C026, C027, C041, C050 ] |
| Name | Type | Why it matters | Evidence |
|---|---|---|---|
| Internal AI builds on customer clouds and tools | Internal Build | Organizations with engineering capacity can use APIs, cloud channels, and their own governance stack to build bespoke assistants or agents instead of buying OpenAI’s packaged SaaS products. | [ C009, C013, C031, C033 ] |
| Manual knowledge-work, support, and operations processes | Manual Process | Customer proofs highlight that OpenAI often competes with repetitive human workflows in research, support, order entry, and operational coordination rather than only with other AI vendors. | [ C023, C028, C029, C030 ] |
| Existing enterprise software and cloud-platform features | Platform Feature | Because OpenAI’s enterprise proposition emphasizes connectors, shared workspaces, admin controls, and deployment inside existing procurement or cloud environments, buyers can compare it with features added by tools they already own. | [ C013, C019, C031, C033, C050 ] |
| Services-led deployment and change-management programs | Services | Rollout success depends on governance, training, and adoption support, so some organizations may address the problem through services and workflow redesign before standardizing on a product platform. | [ C013, C025, C030 ] |
No audited public segment revenue split across consumer ChatGPT, enterprise, API, and government surfaces.
Public evidence on gross margin, burn multiple, customer concentration, and retention remains thin relative to OpenAI’s scale claims.
Non-U.S. public-sector procurement and revenue are much less visible than U.S. government adoption evidence.
Current public disclosure still does not provide a clean operating-org chart or total employee count.
High: Audited segment revenue, gross margin, retention, and customer concentration across consumer, enterprise, API, and government remain undisclosed.
High: Independent validation of the largest 2026 scale claims on users, subscribers, revenue, and valuation is still limited.
Medium: Current public evidence does not cleanly map day-to-day executive ownership across research, products, infrastructure, safety, government, and Foundation activities.
High: Compute economics, cloud commitments, and capex intensity by product line remain too opaque for strong unit-economics underwriting.
Medium: Non-U.S. public-sector procurement scale and recurring revenue contribution remain much less visible than U.S. evidence.
| ID | Statement | Topic | Type | Confidence | Linked sources |
|---|---|---|---|---|---|
| C001 | OpenAI describes itself as an AI research and deployment company. | Identity | Observed | High | S001 |
| C002 | OpenAI currently operates through the nonprofit OpenAI Foundation and the for-profit OpenAI Group PBC, with the Foundation governing the Group. | Governance | Observed | High | S001 S002 S016 |
| C003 | The OpenAI Foundation appoints all members of OpenAI Group’s board and can replace directors at any time. | Governance | Observed | High | S002 |
| C004 | As of the recapitalization close, the OpenAI Foundation held 26% of OpenAI Group, Microsoft held roughly 27%, and current or former employees and investors held the remaining 47%. | Valuation | Observed | High | S002 |
| C005 | OpenAI retains unusual mission-first governance for a commercial AI platform, with the Charter stating that its primary fiduciary duty is to humanity. | Governance | Observed | High | S002 S003 |
| C006 | OpenAI’s privacy policy identifies OpenAI Ireland Limited as controller for EEA and Switzerland users and OpenAI OpCo, LLC for users elsewhere. | Legal | Observed | High | S004 |
| C007 | OpenAI monetizes via consumer subscriptions, business and enterprise workspace seats, and usage-based API pricing. | Business Model | Observed | High | S009 S011 S012 S013 |
| C008 | ChatGPT’s public pricing page lists self-serve consumer tiers from Free to Pro, a Business plan at SGD 25 per user per month billed annually, and Enterprise as custom priced. | Business Model | Observed | High | S011 |
| C009 | OpenAI’s public API pricing is usage-based, with GPT-5.5 listed at $5 per 1M input tokens and $30 per 1M output tokens and Batch API discounts of 50%. | Business Model | Observed | High | S009 |
| C010 | OpenAI’s business offerings say customer business data is not used to train models by default and is protected with encryption, retention controls, and enterprise identity features. | Risk | Observed | High | S005 S006 S012 S013 S014 S015 |
| C011 | OpenAI’s current hiring footprint remains unusually large, with 669 open roles spanning datacenter design, safety, investigations, government, forward-deployed engineering, monetization, and finance. | Team | Observed | High | S010 |
| C012 | OpenAI’s product surface spans consumer ChatGPT, business workspaces, education, government variants, developer APIs, and coding or agent products rather than a single chatbot. | Market | Observed | High | S001 S009 S011 S012 S013 S014 S018 S019 S025 |
| C013 | OpenAI’s enterprise proposition centers on internal-data connectors, shared workspaces, agentic workflows, admin controls, and adoption support rather than standalone model access. | Gtm | Observed | High | S012 S013 S015 S020 |
| C014 | OpenAI explicitly packages sector solutions for financial services and government, indicating a go-to-market strategy around regulated buyers with centralized IT, security, and procurement stakeholders. | Gtm | Observed | High | S019 S020 |
| C015 | OpenAI’s education offering emphasizes campus-wide deployment, no-training defaults, SAML or SCIM, admin controls, and custom retention windows. | Customer | Observed | High | S014 S015 |
| C016 | OpenAI’s ChatGPT Gov announcement says more than 90,000 users across more than 3,500 U.S. government agencies have sent over 18 million messages on ChatGPT since 2024. | Traction | Company Claimed | Medium | S018 |
| C017 | OpenAI’s government solutions page says over 1 million public-sector professionals rely on OpenAI products and highlights a $1 OneGov offer, FedRAMP 20x Moderate, and ChatGPT Gov or AWS GovCloud deployment paths. | Customer | Company Claimed | Medium | S019 |
| C018 | GSA independently announced a OneGov agreement giving participating U.S. federal agencies access to ChatGPT Enterprise for $1 for one year through the Multiple Award Schedule. | Gtm | Third Party Reported | High | S034 |
| C019 | Government adoption depends on procurement and compliance pathways such as GSA, FedRAMP, agency cloud environments, and usage-policy constraints, not just model quality. | Market | Inferred | High | S008 S018 S019 S034 |
| C020 | OpenAI’s trust disclosures show ongoing moderation and legal-process burden, including 224 non-content requests, 75 content requests, 10 emergency requests, and 107,817 NCMEC reports in July to December 2025. | Risk | Observed | High | S007 |
| C021 | OpenAI’s usage policies materially narrow certain addressable use cases by banning or conditioning licensed advice, sensitive decision automation without human review, and national-security or intelligence uses without approval. | Risk | Observed | High | S008 |
| C022 | OpenAI says ChatGPT Enterprise serves more than 5 million business users across industries. | Traction | Company Claimed | Medium | S013 |
| C023 | Customer evidence indicates OpenAI has meaningful deployment beyond pilots: Morgan Stanley reports over 98% adoption among advisor teams, CyberAgent reports 93% monthly active use, and Choco and Gradient Labs cite operational automation outcomes. | Customer | Third Party Reported | Medium | S021 S022 S023 S024 S035 S036 |
| C024 | Morgan Stanley says OpenAI’s zero data retention policy helped address security concerns in wealth-management deployments. | Risk | Company Claimed | Medium | S024 |
| C025 | CyberAgent attributes broad enterprise adoption to governance, training, and admin visibility rather than blanket mandates, implying change management is central to rollout success. | Gtm | Inferred | Medium | S023 |
| C026 | Gradient Labs says GPT-4.1 hit 97% trajectory accuracy versus 88% for the next-best provider in its internal evals, reflecting the importance of reliability in regulated AI support workflows. | Product | Company Claimed | Medium | S022 |
| C027 | Gradient Labs’ own website corroborates a finance-specific AI support positioning with up to 98% CSAT and 40 to 60% day-one resolution rates. | Customer | Third Party Reported | Medium | S035 |
| C028 | Choco says OpenAI-powered systems process 8.8M or more orders annually, 200B or more tokens, cut manual order entry by up to 50%, and double sales productivity without added headcount. | Traction | Company Claimed | Medium | S021 |
| C029 | Choco’s own VoiceAgent announcement corroborates production use of OpenAI’s Realtime API for multilingual phone ordering with ERP integration. | Customer | Third Party Reported | Medium | S036 |
| C030 | Demand is moving from prompting toward agentic execution, internal-data access, and workflow automation across research, coding, support, and operations. | Market | Inferred | High | S012 S013 S020 S021 S022 S023 S024 S025 S031 |
| C031 | OpenAI and AWS now distribute OpenAI models, Codex, and managed agents inside AWS environments, showing OpenAI’s enterprise reach is no longer limited to Microsoft channels. | Competition | Observed | High | S025 S031 |
| C032 | Microsoft’s 2023 partnership announcement made Azure OpenAI’s exclusive cloud provider, highlighting historical platform concentration. | Competition | Third Party Reported | High | S030 |
| C033 | OpenAI’s 2026 Microsoft update says Microsoft remains the primary cloud partner, but OpenAI can now serve all products across any cloud provider and Microsoft’s OpenAI license is non-exclusive. | Competition | Company Claimed | Medium | S026 |
| C034 | The CMA scrutinized whether the Microsoft/OpenAI relationship created a relevant merger situation because of the firms’ close investment, technology, and cloud relationship. | Risk | Third Party Reported | High | S032 S033 |
| C035 | The CMA ultimately decided in March 2025 that Microsoft’s partnership with OpenAI did not qualify for investigation under UK merger provisions. | Risk | Third Party Reported | High | S032 |
| C036 | OpenAI’s safety governance evolved materially in 2024, with the board creating a Safety and Security Committee and later converting it into an independent board oversight committee chaired by Zico Kolter with authority to delay launches. | Governance | Observed | High | S002 S028 S029 |
| C037 | OpenAI Foundation materials suggest a growing philanthropic and resilience agenda alongside the commercial platform, including a $25B initial commitment, a $50M People-First AI Fund, and plans to invest at least $1B over the following year. | Funding | Company Claimed | Medium | S016 S017 |
| C038 | OpenAI’s March 2026 financing post claims $122B in committed capital at an $852B post-money valuation, about $2B in monthly revenue, more than 900M weekly users, over 50M subscribers, and enterprise revenue above 40% of total. | Funding | Company Claimed | Low | S027 |
| C039 | OpenAI does not publicly disclose audited segment revenue, retention, gross margin, burn multiple, or customer concentration across consumer, enterprise, API, and government surfaces. | Financials | Open Question | Low | S011 S013 S027 |
| C040 | OpenAI’s careers page and product surfaces indicate go-to-market expansion into public sector, financial services, life sciences, startups, ads, and enterprise deployment geographies across North America, Europe, and APAC. | Gtm | Inferred | High | S010 S019 S020 |
| C041 | OpenAI’s broad market coverage gives it multiple budget entry points, including individual discretionary spend, engineering and product budgets, central IT or digital-transformation budgets, and government procurement vehicles. | Market | Inferred | High | S009 S011 S012 S013 S018 S019 S020 S034 |
| C042 | OpenAI is using official security, privacy, and compliance artifacts to lower enterprise adoption friction, including SOC 2, ISO 27001 family certifications, ISO 42001, data residency, DPA or BAA support, and audit visibility. | Risk | Observed | High | S005 S006 S013 S015 |
| C043 | OpenAI’s 2015 founding as a nonprofit mission-driven venture and its 2019 shift to a capped-profit structure are independently corroborated by BBC and TechCrunch. | Identity | Third Party Reported | High | S037 S038 |
| C044 | TechCrunch reported that OpenAI’s 2019 structural shift was driven by the need to invest billions in compute, talent, and AI supercomputers, underscoring persistent capital intensity in frontier AI. | Funding | Third Party Reported | Medium | S038 |
| C045 | OpenAI faces key-person and leadership-continuity risk: CNBC reported the 2024 departures of Mira Murati, Bob McGrew, and Barret Zoph, while TechCrunch reported John Schulman’s exit and that only three of the 11 original founders remained active. | Team | Third Party Reported | High | S039 S040 |
| C046 | OpenAI’s safety and moderation obligations, regulatory scrutiny, and leadership turnover together suggest that operational complexity is scaling alongside revenue opportunity. | Risk | Inferred | Medium | S007 S028 S032 S039 S040 |
| C047 | Public evidence remains far richer for U.S. public-sector adoption than for non-U.S. government revenue or procurement scale. | Market | Open Question | Low | S018 S019 S034 |
| C048 | OpenAI’s market appears attractive because it combines strong consumer pull, enterprise workflow expansion, developer spend, and emerging public-sector procurement, but the public unit-economics picture remains incomplete. | Market | Inferred | Medium | S012 S013 S018 S019 S025 S027 |
| C049 | Foundation-side hiring and program announcements indicate OpenAI is expanding governance-adjacent and mission-side staffing in parallel with commercial scaling. | Team | Observed | High | S017 |
| C050 | Customer-side and partner-side evidence show OpenAI’s distribution now runs through direct products, AWS Bedrock, Azure or Azure Government, and GSA procurement channels. | Gtm | Third Party Reported | High | S030 S031 S034 |
| C051 | OpenAI’s 2019 capped-profit structure gave the nonprofit control, put the Charter ahead of investor economics, limited conflicted board voting, and capped early investor returns at 100x. | Governance | Observed | High | S041 |
| C052 | CNBC reported that OpenAI closed a $40B March 2025 financing at a $300B post-money valuation, led by $30B from SoftBank and $10B from a syndicate including Microsoft, Coatue, Altimeter, and Thrive. | Funding | Third Party Reported | High | S043 |
| C053 | CNBC reported that about $18B of the March 2025 funding was expected to support Stargate and that SoftBank’s total investment could be reduced if OpenAI did not complete a for-profit restructuring by year-end; venture backers received convertible notes that would turn into equity. | Governance | Third Party Reported | Medium | S043 |
| C054 | CNBC reported that OpenAI reached 3M paying business users in June 2025, up from 2M in February, across ChatGPT Enterprise, Team, and Edu. | Traction | Third Party Reported | High | S044 |
| C055 | CNBC reported that OpenAI reached $10B in annual recurring revenue in June 2025; the figure included consumer products, ChatGPT business products, and API sales, and excluded Microsoft licensing revenue and large one-time deals. | Financials | Third Party Reported | High | S045 |
| C056 | CNBC reported that OpenAI had about $5.5B in ARR for 2024, lost about $5B that year, and that the March 2025 $300B valuation equated to about 30x revenue at June 2025 metrics. | Valuation | Third Party Reported | Medium | S045 |
| C057 | OpenAI’s 2026 Microsoft update says Microsoft remains the primary cloud partner, OpenAI can serve products across any cloud, Microsoft has a non-exclusive IP license through 2032, Microsoft no longer pays revenue share to OpenAI, and OpenAI revenue-share payments to Microsoft continue through 2030 subject to a cap. | Financials | Company Claimed | Medium | S053 |
| C058 | SoftBank disclosed a planned $30B 2026 follow-on investment in OpenAI Group PBC at a $730B pre-money valuation via automatically converting preferred shares, with cumulative OpenAI investment expected to reach $64.6B for about 13% ownership. | Funding | Third Party Reported | High | S046 |
| C059 | SoftBank disclosed a $40B unsecured bridge facility for the OpenAI follow-on investment and general corporate purposes, then executed the first $10B tranche on April 1, 2026. | Funding | Third Party Reported | High | S047 S048 |
| C060 | OpenAI announced Stargate as a $500B, four-year AI-infrastructure effort with $100B to deploy immediately; it later said Oracle expansion brought Stargate capacity under development to over 5GW and over 2M chips. | Technology | Company Claimed | Medium | S042 S049 |
| C061 | CoreWeave’s SEC prospectus reported 2024 revenue of $1.9B, $15.1B of RPO, 32 active data centers, more than 250,000 GPUs, approximately 1.3GW contracted power, 2024 net loss of $863M, and cost of revenue equal to 26% of revenue. | Valuation | Observed | High | S050 |
| C062 | CoreWeave’s SEC prospectus disclosed an OpenAI Master Services Agreement under which OpenAI committed to pay up to approximately $11.9B through October 2030, plus a $350M concurrent CoreWeave share issuance to OpenAI tied to the IPO. | Financials | Observed | High | S050 |
| C063 | Anthropic disclosed a $13B Series F at a $183B post-money valuation, over $5B run-rate revenue by August 2025, over 300,000 business customers, and large-account growth of nearly 7x year over year. | Valuation | Company Claimed | Medium | S051 |
| C064 | Anthropic disclosed a more than $100B, ten-year AWS technology commitment securing up to 5GW of new capacity, alongside Amazon investing $5B immediately and up to $20B more in the future. | Technology | Company Claimed | Medium | S052 |
| ID | Publisher | Title | Independence | Quote |
|---|---|---|---|---|
| S001 | OpenAI | About | Company | OpenAI is an AI research and deployment company. Our mission is to ensure that artificial general intelligence benefits all of humanity. |
| S002 | OpenAI | Our structure | Company | OpenAI consists of the nonprofit OpenAI Foundation and the for-profit OpenAI Group. The Foundation governs the Group, which operates as a public benefit corporation. |
| S003 | OpenAI | OpenAI Charter | Company | Our primary fiduciary duty is to humanity. |
| S004 | OpenAI | Privacy policy | Company | If you live anywhere else, OpenAI OpCo, LLC, with its registered office at 1455 Third Street, San Francisco, California 94158, United States, is the controller and is responsible for the processing of your Personal Data as described in this policy. |
| S005 | OpenAI | Business data privacy, security, and compliance | Company | By default, we do not use data from ChatGPT Enterprise, ChatGPT Business, ChatGPT Edu, ChatGPT for Healthcare, ChatGPT for Teachers, or our API platform—including inputs or outputs—for training or improving our models. |
| S006 | OpenAI | Security & privacy | Company | OpenAI has undergone an independent SOC 2 Type 2 examination of controls relevant to Security, Availability, Confidentiality, and Privacy for its API and ChatGPT business product services. |
| S007 | OpenAI | Trust & transparency | Company | 224 Non-Content Requests July–December 2025 |
| S008 | OpenAI | Usage policies | Company | provision of tailored advice that requires a license, such as legal or medical advice, without appropriate involvement by a licensed professional |
| S009 | OpenAI | API Pricing | Company | Input:$5.00 / 1M tokensCached input:$0.50 / 1M tokensOutput:$30.00 / 1M tokens |
| S010 | OpenAI | Careers at OpenAI | Company | 669 jobs |
| S011 | OpenAI | Pricing | Company | SGD 25 / user / month |
| S012 | OpenAI | ChatGPT for business, powered by OpenAI’s most advanced models | Company | ChatGPT Business is the fastest way to start using ChatGPT for work—with a shared workspace, admin controls, and apps for your company tools. |
| S013 | OpenAI | Frontier AI built for enterprise | Company | Over 5 million business users across industries |
| S014 | OpenAI | Bring AI to campus at scale | Company | No data or conversations used to train OpenAI models |
| S015 | OpenAI | Enterprise privacy at OpenAI | Company | You own and control your data |
| S016 | OpenAI Foundation | OpenAI Foundation | Company | The OpenAI Foundation made an initial $25 billion commitment across two programs |
| S017 | OpenAI Foundation | Update on the OpenAI Foundation | Company | the Foundation expects to invest at least $1 billion across life sciences and curing diseases, jobs and economic impact, AI resilience, and community programs. |
| S018 | OpenAI | Introducing ChatGPT Gov | Company | Since 2024, more than 90,000 users across more than 3,500 US federal, state, and local government agencies have sent over 18 million messages on ChatGPT |
| S019 | OpenAI | AI that meets the needs of government | Company | Join over 1 million of your U.S. government colleagues using ChatGPT Enterprise for only $1. |
| S020 | OpenAI | Reimagine financial services with AI you can trust | Company | Built for the standards of finance—with the security, reliability, and control enterprises require. |
| S021 | OpenAI | Choco automates food distribution with AI agents | Company | Processes over 8.8 million orders annually, eliminating millions of manual workflows |
| S022 | OpenAI | Gradient Labs gives every bank customer an AI account manager | Company | In one of their initial evals, GPT‑4.1 was the only model to hit 97% trajectory accuracy and consistency. The next closest provider was 88%. |
| S023 | OpenAI | CyberAgent moves faster with ChatGPT Enterprise and Codex | Company | Even so, it is now used across nearly all departments, reaching a monthly active user rate of 93%. |
| S024 | OpenAI | Morgan Stanley uses AI evals to shape the future of financial services | Company | Nearly all advisor teams now use AI tools like the Assistant daily, achieving over 98% adoption in wealth management. |
| S025 | OpenAI | OpenAI models, Codex, and Managed Agents come to AWS | Company | Customers can now build with OpenAI models in AWS, alongside the services, security controls, identity systems, and procurement processes they already rely on. |
| S026 | OpenAI | The next phase of the Microsoft OpenAI partnership | Company | Microsoft remains OpenAI’s primary cloud partner, and OpenAI products will ship first on Azure, unless Microsoft cannot and chooses not to support the necessary capabilities. OpenAI can now serve all its products to customers across any cloud provider. |
| S027 | OpenAI | OpenAI raises $122 billion to accelerate the next phase of AI | Company | Today, we closed our latest funding round with $122 billion in committed capital at a post money valuation of $852 billion. |
| S028 | OpenAI | An update on our safety & security practices | Company | The Safety and Security Committee will become an independent Board oversight committee focused on safety and security |
| S029 | OpenAI | OpenAI board forms safety and security committee | Company | This new committee is responsible for making recommendations on critical safety and security decisions for all OpenAI projects |
| S030 | Microsoft | Microsoft and OpenAI extend partnership | Partner | As OpenAI’s exclusive cloud provider, Azure will power all OpenAI workloads across research, products and API services. |
| S031 | Amazon Web Services | Amazon Bedrock Managed Agents, powered by OpenAI (Limited preview) | Partner | All inference runs on Amazon Bedrock and your data never leaves AWS. |
| S032 | Competition and Markets Authority | Microsoft / OpenAI partnership merger inquiry | Independent | 5 March 2025: The CMA has decided that Microsoft’s partnership with OpenAI does not qualify for investigation under the merger provisions of the Enterprise Act 2002. |
| S033 | Competition and Markets Authority | CMA seeks views on Microsoft’s partnership with OpenAI | Independent | The partnership between Microsoft and OpenAI (including a multi-year, multi-billion dollar investment, collaboration in technology development and exclusive provision of cloud services by Microsoft to OpenAI) represents a close, multi-faceted relationship between two firms with significant activities in FMs and related markets. |
| S034 | U.S. General Services Administration | GSA Announces New Partnership with OpenAI, Delivering Deep Discount to ChatGPT Gov-Wide Through MAS | Independent | Every participating U.S. federal agency will have access for a nominal fee of $1 to ChatGPT Enterprise for one year. |
| S035 | Gradient Labs | The only AI support agent built for financial services | Customer | Our AI Agent delivers 80+ CSAT across all our customers, reaching up to 98% in optimal implementations. |
| S036 | Choco | Introducing Choco VoiceAgent, built by Choco in collaboration with OpenAI | Customer | It has been developed by Choco in collaboration with OpenAI, using OpenAI’s Realtime API. |
| S037 | BBC News | Tech giants pledge $1bn for ''altruistic AI'' venture, OpenAI | Independent | Prominent tech executives have pledged $1bn (£659m) for OpenAI, a non-profit venture that aims to develop artificial intelligence (AI) to benefit humanity. |
| S038 | TechCrunch | OpenAI shifts from nonprofit to ‘capped-profit’ to attract capital | Independent | We’ll need to invest billions of dollars in upcoming years into large-scale cloud compute, attracting and retaining talented people, and building AI supercomputers. |
| S039 | CNBC | OpenAI considering restructuring to for-profit, CTO Mira Murati and two top research execs depart | Independent | OpenAI’s board is considering plans to restructure the firm to a for-profit business |
| S040 | TechCrunch | OpenAI co-founder Schulman leaves for Anthropic, Brockman takes extended leave | Independent | With Schulman’s departure, only three of OpenAI’s 11 original founders remain: OpenAI CEO Sam Altman, Brockman and Wojciech Zaremba |
| S041 | OpenAI | OpenAI LP | Company | OpenAI LP’s primary fiduciary obligation is to advance the aims of the OpenAI Charter, and the company is controlled by OpenAI Nonprofit’s board. |
| S042 | OpenAI | Announcing The Stargate Project | Company | The Stargate Project is a new company which intends to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States. |
| S043 | CNBC | OpenAI closes $40 billion funding round, largest private tech deal on record | Independent | The $40 billion financing values the ChatGPT maker at $300 billion, including the fresh capital. |
| S044 | CNBC | OpenAI tops 3 million paying business users, launches new features for workplace | Independent | OpenAI announced that it now has 3 million paying business users, up from 2 million in February. |
| S045 | CNBC | OpenAI hits $10 billion in annual recurring revenue fueled by ChatGPT growth | Independent | The figure includes sales from the company’s consumer products; ChatGPT business products; and its application programming interface, or API. It excludes licensing revenue from Microsoft and large one-time deals. |
| S046 | SoftBank Group | Follow-on Investments in OpenAI | Investor | Upon completion of the Follow-on Investment, SBG’s cumulative investment in OpenAI is expected to total USD 64.6 billion, representing an ownership interest of approximately 13%. |
| S047 | SoftBank Group | Execution of Bridge Facility Agreement Primarily for the Follow-on Investments in OpenAI | Investor | SoftBank Group Corp. today announced that it entered into a bridge facility agreement with a total facility amount of USD 40.0 billion. |
| S048 | SoftBank Group | Execution of Follow-on Investment (First Tranche) in OpenAI | Investor | SBG today announced that, on April 1, 2026, it executed the follow-on investment (first tranche) of USD 10.0 billion. |
| S049 | OpenAI | Stargate advances with 4.5 GW partnership with Oracle | Company | Together with our Stargate I site in Abilene, Texas, this additional partnership with Oracle will bring us to over 5 gigawatts of Stargate AI data center capacity under development, which will run over 2 million chips. |
| S050 | SEC / CoreWeave | CoreWeave, Inc. 424B4 prospectus | Independent | OpenAI has committed to pay us up to approximately $11.9 billion through October 2030. |
| S051 | Anthropic | Anthropic raises $13B Series F at $183B post-money valuation | Company | Anthropic has completed a Series F fundraising of $13 billion led by ICONIQ. This financing values Anthropic at $183 billion post-money. |
| S052 | Anthropic | Anthropic and Amazon expand collaboration for up to 5 gigawatts of new compute | Company | We are committing more than $100 billion over the next ten years to AWS technologies, securing up to 5GW of new capacity to train and run Claude. |
| S053 | OpenAI | The next phase of the Microsoft OpenAI partnership | Company | Revenue share payments from OpenAI to Microsoft continue through 2030, independent of OpenAI’s technology progress, at the same percentage but subject to a total cap. |