OpenAI is organised as a nonprofit that controls a for-profit company. The nonprofit, OpenAI, Inc. (now called the OpenAI Foundation), was incorporated in Delaware in 2015; since the recapitalization of 28 October 2025 it controls OpenAI Group PBC, a public benefit corporation, and has the sole power to appoint and remove its directors. Investors such as Microsoft, which held an investment of roughly 27 percent valued at about $135 billion after the deal, own equity in the PBC but do not control it.
The Delaware Attorney General's conditions give the nonprofit the sole power to appoint and remove members of the PBC board, so control does not follow ownership.
The PBC's mission is identical to the nonprofit's: to ensure that artificial general intelligence benefits all of humanity, using the principles in the OpenAI Charter.
03
Safety above profit
On safety and security, PBC directors may consider only the mission, not shareholders' financial interests. The Safety and Security Committee, chaired by Zico Kolter, remains a nonprofit committee with observer rights at the PBC board.
04
Investors hold equity
Microsoft's stake was about 27 percent after the recapitalization and about 25 percent on an as-converted basis at 30 June 2026, accounted for under the equity method.
What changed in 2025-26
The October 2025 recapitalization
After a review of more than a year, Delaware Attorney General Kathy Jennings issued a Statement of No Objection on 28 October 2025, in coordination with California’s Attorney General. OpenAI had first proposed removing nonprofit oversight, then kept the nonprofit in control.
A new Microsoft agreement
The same day Microsoft and OpenAI signed a new agreement: Microsoft’s IP rights run through 2032, an independent expert panel will verify any AGI declaration, OpenAI contracted to buy an incremental $250 billion of Azure services, and Microsoft gave up its right of first refusal on compute.
More cloud partners
Under the new terms, API products stay exclusive to Azure but non-API products may run on any cloud provider, and OpenAI can release open-weight models that meet capability criteria.
Money flowing both ways
Microsoft recorded $24.1 billion of revenue from OpenAI, including revenue-sharing payments, in its fiscal year to 30 June 2026, and $6.5 billion of net gains from its OpenAI investment, mainly from the recapitalization.
Who does what in OpenAI's structure
OpenAI Foundation (nonprofit)
RoleControls the PBC; appoints and removes its directors
SourceDelaware AG, 28 October 2025
Safety and Security Committee
RoleOversees safety; can require mitigations, up to halting releases
SourceDelaware AG, 28 October 2025
OpenAI Group PBC
RoleOperating company; mission identical to the nonprofit's
SourceDelaware AG, 28 October 2025
Microsoft
RoleAbout 25 percent investor; IP rights through 2032; Azure API exclusivity
SourceMicrosoft 10-K (fiscal 2026) and Microsoft blog
What it means for you
For investors: owning OpenAI shares does not give control; the nonprofit appoints the board.
For enterprise buyers: OpenAI API products remain on Azure, while other products may run on other clouds.
For policy watchers: safety decisions sit with a nonprofit committee that can stop a model release.
For founders: the structure shows a way to raise very large sums while keeping a mission lock, at the cost of complexity.
Update old references: the 2019 capped-profit OpenAI LP structure has been replaced by the PBC.
The Business Engineer lens
OpenAI’s structure separates who owns the company from who controls it, which lets it raise capital at very large scale while a nonprofit keeps the mission and safety veto. In the AI economy this matters because compute needs force frontier labs into deep ties with hyperscalers, and governance decides who captures the value.
A nonprofit, the OpenAI Foundation, controls OpenAI Group PBC, a public benefit corporation, and appoints and removes its board.
Is OpenAI still a nonprofit?
The parent is a Delaware nonprofit; the operating business is a for-profit public benefit corporation that the nonprofit controls.
How much of OpenAI does Microsoft own?
About 27 percent, valued at about $135 billion, after the October 2025 recapitalization; about 25 percent on an as-converted basis as of 30 June 2026, per Microsoft's 10-K.
Who decides on safety at OpenAI?
The nonprofit's Safety and Security Committee, which can require mitigations up to halting a model release.
The complete explainer, cleaned up and reformatted for mobile.
OpenAI is an artificial intelligence research laboratory that created a capped-profit arm in 2019. The corporate structure is organized around two entities: OpenAI, Inc., the Delaware nonprofit, and OpenAI LP, a capped-profit company it controlled until the October 2025 recapitalization into OpenAI Group PBC. The OpenAI LP is governed by the board of OpenAI, Inc (the foundation), which acts as a General Partner. At the same time, Limited Partners comprise employees of the LP, some of the board members, and other investors like Reid Hoffman’s charitable foundation, Khosla Ventures, and Microsoft, which, according to Microsoft in October 2025, holds an investment in OpenAI Group PBC valued at approximately $135 billion, representing roughly 27 percent on an as-converted diluted basis.
Updated 6 October 2026: we refreshed 7 figures on this page, now taken from blogs.microsoft.com, convergedigest.com and siliconangle.com. Anything we could not check against a current source keeps its original year, so you can see how old it is.
Latest figures · checked 6 October 2026
In October 2025, the OpenAI Foundation committed an initial $25 billion to health research and AI resilience technologies, according to Converge Digest. convergedigest.com
As of October 2025, Microsoft’s investment in OpenAI Group PBC was valued at approximately $135 billion, roughly 27 percent on an as-converted diluted basis. blogs.microsoft.com
Under the October 2025 agreement, Microsoft’s IP rights to OpenAI models and products are extended through 2032. blogs.microsoft.com
In March 2026, OpenAI Group PBC announced it had raised $122 billion at an $852 billion valuation, according to SiliconANGLE. siliconangle.com
As of March 2026, OpenAI said it was generating in excess of $2 billion in sales per month, according to SiliconANGLE. siliconangle.com
As of March 2026, OpenAI said it had more than 900 million weekly active users of its consumer applications, according to SiliconANGLE. siliconangle.com
History of OpenAI
Key numbersChecked 6 October 2026
$25 billionIn October 2025, the OpenAI Foundation committed an initial $25 billion to health research and AI…convergedigest.com
$135 billionAs of October 2025, Microsoft's investment in OpenAI Group PBC was valued at approximately $135 billion…blogs.microsoft.com
$122 billionIn March 2026, OpenAI Group PBC announced it had raised $122 billion at an $852 billion valuation…siliconangle.com
$2 billionAs of March 2026, OpenAI said it was generating in excess of $2 billion in sales per month, according to…siliconangle.com
In one lineOpenAI is an artificial intelligence research laboratory that created a capped-profit arm in 2019.
FourWeekMBABusiness EngineerAt a glance
As of October 2025, Microsoft's investment in OpenAI Group PBC was valued at approximately $135 billion, roughly 27 percent on an as-converted diluted basis. blogs.microsoft.com
In October 2025, the OpenAI Foundation committed an initial $25 billion to health research and AI resilience technologies, according to Converge Digest. convergedigest.com
The team operates with direct reporting lines to leadership and veto power over model releases, fundamentally altering decision-making hierarchies within the company.
Layout and figures refreshed 6 October 2026. Sources are listed at the end of the article.
OpenAI Business Model
FourWeekMBABusiness EngineerTimeline, as stated in the article
2019
OpenAI is an artificial intelligence research laboratory that created a capped-profit arm in 2019.
2025
In October 2025, the OpenAI Foundation committed an initial $25 billion to health research and AI resilience technologies, according to Converge Digest. convergedigest.com
2025
As of October 2025, Microsoft's investment in OpenAI Group PBC was valued at approximately $135 billion, roughly 27 percent on an as-converted diluted basis. blogs.microsoft.com
2025
Under the October 2025 agreement, Microsoft's IP rights to OpenAI models and products are extended through 2032. blogs.microsoft.com
2026
In March 2026, OpenAI Group PBC announced it had raised $122 billion at an $852 billion valuation, according to SiliconANGLE. siliconangle.com
2026
As of March 2026, OpenAI said it had more than 900 million weekly active users of its consumer applications, according to SiliconANGLE. siliconangle.com
OpenAI’s mission
OpenAI’s missionis to ensure that artificial general intelligence (AGI).
As OpenAI explains, “by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity.”
How does OpenAI corporate structure work?
OpenAI is broken down in:
OpenAI LP, a capped-profit entity (replaced by OpenAI Group PBC in October 2025).
OpenAI Inc. board is the general partner, which leads, managed, and controls the decision made by the OpenAI LP.
On the other hand, limited partners, which had no control over the management of the LP, also had a capped return of 100x on their investments.
The cap is based on agreements that are in place for each of the limited partners.
Microsoft commercial partnership
Microsoft was the leading limited partner, and it also had a commercial agreement as part of the one billion invested in 2019, which comprises the following:
Microsoft/OpenAI Azure AI supercomputing technologies.
OpenAI products are served through Microsoft Azure.
Microsoft is OpenAI’s preferred partner for commercializing new AI technologies.
In the meantime, a further partnership was closed on January 23rd, 2023, where Microsoft and OpenAI confirmed a multi-year, multi-billion agreement where Microsoft announced three main parts of this partnership:
Supercomputing at scale – Microsoft will increase our investments in the development and deployment of specialized supercomputing systems to accelerate OpenAI’s groundbreaking independent AI research. We will also continue to build out Azure’s leading AI infrastructure to help customers build and deploy their AI applications on a global scale.
New AI-powered experiences – Microsoft will deploy OpenAI’s models across our consumer and enterprise products and introduce new categories of digital experiences built on OpenAI’s technology. This includes Microsoft’s Azure OpenAI Service, which empowers developers to build cutting-edge AI applications through direct access to OpenAI models backed by Azure’s trusted, enterprise-grade capabilities and AI-optimized infrastructure and tools.
Exclusive cloud provider – As OpenAI’s exclusive cloud provider, Azure will power all OpenAI workloads across research, products and API services.
Key Highlights
History of OpenAI:
OpenAI started as an artificial intelligence research laboratory.
In 2019, it transitioned into a for-profit organization.
The organization’s structure includes two main entities: OpenAI, Inc. and OpenAI LP.
OpenAI’s Mission:
OpenAI’s mission is to ensure the development of artificial general intelligence (AGI) that benefits humanity.
AGI refers to highly autonomous systems that surpass human performance in economically valuable tasks.
Corporate Structure:
OpenAI LP was a capped-profit entity until the October 2025 recapitalization.
The board of OpenAI, Inc. serves as the General Partner for OpenAI LP.
Limited Partners, including employees, board members, investors like Microsoft, had capped returns and limited control.
Microsoft Partnership:
Microsoft was a leading limited partner in OpenAI before the 2025 recapitalization.
Microsoft invested a billion dollars in 2019 and entered into a commercial agreement.
The partnership involves various aspects:
Microsoft contributes to OpenAI’s supercomputing technologies.
OpenAI products are hosted on Microsoft Azure.
Microsoft deploys OpenAI models across its products, introducing new AI-powered experiences.
Azure was OpenAI’s exclusive cloud provider under the 2023 agreement; since October 2025 only API products are exclusive to Azure.
Recent Multi-Year Agreement:
In January 2023, Microsoft and OpenAI announced a multi-year, multi-billion agreement.
The agreement includes:
Microsoft’s increased investments in specialized supercomputing systems to aid OpenAI’s AI research.
Deployment of OpenAI’s models in Microsoft’s consumer and enterprise products.
Microsoft Azure as the exclusive cloud provider for OpenAI’s workloads.
Large language models (LLMs) are AI tools that can read, summarize, and translate text. This enables them to predict words and craft sentences that reflect how humans write and speak.
Prompt engineering is a natural language processing (NLP) concept that involves discovering inputs that yield desirable or useful results.
Like most processes, the quality of the inputs determines the quality of the outputs in prompt engineering. Designing effective prompts increases the likelihood that the model will return a response that is both favorable and contextual.
Developed by OpenAI, the CLIP (Contrastive Language-Image Pre-training) model is an example of a model that utilizes prompts to classify images and captions from over 400 million image-caption pairs.
OpenAI has built the foundational layer of the AI industry. With large generative models like GPT-3 and DALL-E, OpenAI offers API access to businesses that want to develop applications on top of its foundational models while being able to plug these models into their products and customize these models with proprietary data and additional AI features. On the other hand, OpenAI also released ChatGPT, developing around a freemium model. Microsoft also commercializes opener products through its commercial partnership.
OpenAI and Microsoft partnered up from a commercial standpoint. The history of the partnership started in 2016 and consolidated in 2019, with Microsoft investing a billion dollars into the partnership. It’s now taking a leap forward, with Microsoft in talks to put $10 billion into this partnership. Microsoft, through OpenAI, is developing its Azure AI Supercomputer while enhancing its Azure Enterprise Platform and integrating OpenAI’s models into its business and consumer products (GitHub, Office, Bing).
Stability AI is the entity behind Stable Diffusion. Stability makes money from our AI products and from providing AI consulting services to businesses. Stability AI monetizes Stable Diffusion via DreamStudio’s APIs. While it also releases it open-source for anyone to download and use. Stability AI also makes money via enterprise services, where its core development team offers the chance to enterprise customers to service, scale, and customize Stable Diffusion or other large generative models to their needs.
OpenAI is an artificial intelligence research laboratory that created a capped-profit arm in 2019. The corporate structure is organized around two entities: OpenAI, Inc., the Delaware nonprofit, and OpenAI LP, a capped-profit company it controlled until the October 2025 recapitalization into OpenAI Group PBC. The OpenAI LP is governed by the board of OpenAI, Inc (the foundation), which acts as a General Partner. At the same time, Limited Partners comprise employees of the LP, some of the board members, and other investors like Reid Hoffman’s charitable foundation, Khosla Ventures, and Microsoft, which, according to Microsoft in October 2025, holds an investment in OpenAI Group PBC valued at approximately $135 billion, representing roughly 27 percent on an as-converted diluted basis.
Airbnb follows a holacracy model, or a sort of flat organizational structure, where teams are organized for projects, to move quickly and iterate fast, thus keeping a lean and flexible approach. Airbnb also moved to a hybrid model where employees can work from anywhere and meet on a quarterly basis to plan ahead, and connect to each other.
The Amazon organizational structure is predominantly hierarchical with elements of function-based structure and geographic divisions. While Amazon started as a lean, flat organization in its early years, it transitioned into a hierarchical organization with its jobs and functions clearly defined as it scaled.
The Coca-Cola Company has a somewhat complex matrix organizational structure with geographic divisions, product divisions, business-type units, and functional groups.
Costco has a matrix organizational structure, which can simply be defined as any structure that combines two or more different types. In this case, a predominant functional structure exists with a more secondary divisional structure.
Costco’s geographic divisions reflect its strong presence in the United States combined with its expanding global presence. There are six divisions in the country alone to reflect its standing as the source of most company revenue.
Compared to competitor Walmart, for example, Costco takes more a decentralized approach to management, decision-making, and autonomy. This allows the company’s stores and divisions to more flexibly respond to local market conditions.
Dell has a functional organizational structure with some degree of decentralization. This means functional departments share information, contribute ideas to the success of the organization and have some degree of decision-making power.
eBay was until recently a multi-divisional (M-form) organization with semi-autonomous units grouped according to the services they provided. Today, eBay has a single division called Marketplace, which includes eBay and its international iterations.
Facebook is characterized by a multi-faceted matrix organizational structure. The company utilizes a flat organizational structure in combination with corporate function-based teams and product-based or geographic divisions. The flat organization structure is organized around the leadership of Mark Zuckerberg, and the key executives around him. On the other hand, the function-based teams are based on the main corporate functions (like HR, product management, investor relations, and so on).
Goldman Sachs has a hierarchical structure with a clear chain of command and defined career advancement process. The structure is also underpinned by business-type divisions and function-based groups.
Google (Alphabet) has a cross-functional (team-based) organizational structure known as a matrix structure with some degree of flatness. Over the years, as the company scaled and it became a tech giant, its organizational structure is morphing more into a centralized organization.
IBM has an organizational structure characterized by product-based divisions, enabling its strategy to develop innovative and competitive products in multiple markets. IBM is also characterized by function-based segments that support product development and innovation for each product-based division, which include Global Markets, Integrated Supply Chain, Research, Development, and Intellectual Property.
McDonald’s has a divisional organizational structure where each division – based on geographical location – is assigned operational responsibilities and strategic objectives. The main geographical divisions are the US, internationally operated markets, and international developmental licensed markets. And on the other hand, the hierarchical leadership structure is organized around regional and functional divisions.
McKinsey & Company has a decentralized organizational structure with mostly self-managing offices, committees, and employees. There are also functional groups and geographic divisions with proprietary names.
Microsoft has a product-type divisional organizational structure based on functions and engineering groups. As the company scaled over time it also became more hierarchical, however still keeping its hybrid approach between functions, engineering groups, and management.
Nestlé has a geographical divisional structure with operations segmented into five key regions. For many years, Swiss multinational food and drink company Nestlé had a complex and decentralized matrix organizational structure where its numerous brands and subsidiaries were free to operate autonomously.
Nike has a matrix organizational structure incorporating geographic divisions. Nike’s matrix structure is also present at the regional and sub-regional levels. Managerial responsibility is segmented according to business unit (apparel, footwear, and equipment) and function (human resources, finance, marketing, sales, and operations).
Patagonia has a particular organizational structure, where its founder, Chouinard, disposed of the company’s ownership in the hands of two non-profits. The Patagonia Purpose Trust, holding 100% of the voting stocks, is in charge of defining the company’s strategic direction. And the Holdfast Collective, a non-profit, holds 100% of non-voting stocks, aiming to re-invest the brand’s dividends into environmental causes.
Samsung has a product-type divisional organizational structure where products determine how resources and business operations are categorized. The main resources around which Samsung’s corporate structure is organized are consumer electronics, IT, and device solutions. In addition, Samsung leadership functions are organized around a few career levels grades, based on experience (assistant, professional, senior professional, and principal professional).
Sony has a matrix organizational structure primarily based on function-based groups and product/business divisions. The structure also incorporates geographical divisions. In 2021, Sony announced the overhauling of its organizational structure, changing its name from Sony Corporation to Sony Group Corporation to better identify itself as the headquarters of the Sony group of companies skewing the company toward product divisions.
Starbucks follows a matrix organizational structure with a combination of vertical and horizontal structures. It is characterized by multiple, overlapping chains of command and divisions.
Tesla is characterized by a functional organizational structure with aspects of a hierarchical structure. Tesla does employ functional centers that cover all business activities, including finance, sales, marketing, technology, engineering, design, and the offices of the CEO and chairperson. Tesla’s headquarters in Austin, Texas, decide the strategic direction of the company, with international operations given little autonomy.
Toyota has a divisional organizational structure where business operations are centered around the market, product, and geographic groups. Therefore, Toyota organizes its corporate structure around global hierarchies (most strategic decisions come from Japan’s headquarter), product-based divisions (where the organization is broken down, based on each product line), and geographical divisions (according to the geographical areas under management).
Walmart has a hybrid hierarchical-functional organizational structure, otherwise referred to as a matrix structure that combines multiple approaches. On the one hand, Walmart follows a hierarchical structure, where the CEO (John Furner since 1 February 2026) is the only employee without a direct superior, and directives are sent from top-level management. On the other hand, the function-based structure of Walmart is used to categorize employees according to their particular skills and experience.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
