OpenAI

OpenAI: An Organization Navigating the Frontier of Artificial Intelligence

OpenAI stands as a preeminent force in the field of artificial intelligence (AI) research and development. Headquartered in the bustling innovation hub of San Francisco, California , the organization’s core mission is the ambitious pursuit of “safe and beneficial” artificial general intelligence (AGI). They define AGI as systems exhibiting a high degree of autonomy, capable of surpassing human performance across the majority of economically valuable tasks. As a central player in the current AI boom , OpenAI has garnered significant recognition for its groundbreaking contributions, including the widely influential GPT series of large language models , the visually stunning DALL-E family of text-to-image models , and the innovative text-to-video model known as Sora . The public release of ChatGPT in November 2022, in particular, is credited with igniting a widespread surge of interest and development in the realm of generative AI .

While initially founded in Delaware on December 8, 2015, OpenAI has since evolved into a complex corporate entity. As of October 2025, a significant restructuring, approved by regulators in both California and Delaware, established a new framework. The non-profit OpenAI Foundation now holds a 26% stake in the for-profit OpenAI Group PBC, a public benefit corporation . Microsoft , a crucial strategic partner, holds a 27% stake, with employees and other investors collectively owning the remaining 47%. The OpenAI Foundation retains ultimate control, appointing all members of the for-profit entity’s board. Microsoft’s commitment to OpenAI is substantial, having invested over $13 billion to date and providing critical Azure cloud computing resources. The company’s financial trajectory has been marked by significant growth, with reported revenues of US$3.7 billion in 2024, though it also incurred a net loss of US$5 billion in the same year. As of 2025, OpenAI employed approximately 3,000 individuals.

The organization has not been without its challenges and controversies. In 2023 and 2024, OpenAI faced numerous lawsuits alleging copyright infringement by creators and media entities whose works were allegedly used to train its products. A notable internal upheaval occurred in November 2023 when the board of directors controversially removed Sam Altman as CEO, citing a loss of confidence. However, following intense pressure and negotiations, Altman was reinstated just five days later, accompanied by a reconstitution of the board. Throughout 2024, a significant exodus of AI safety researchers, approximately half of the department, occurred, with departing employees citing the company’s central role in industry-wide concerns.

Founding and Early Vision

OpenAI was established in December 2015 by a group of prominent figures in the technology and AI landscape, including Sam Altman , Elon Musk , Ilya Sutskever , Greg Brockman , Trevor Blackwell , Vicki Cheung, Andrej Karpathy , Durk Kingma, John Schulman , Pamela Vagata, and Wojciech Zaremba . The initial capital pledges, totaling $1 billion, came from a diverse group of individuals and entities such as Sam Altman, Greg Brockman, Elon Musk, Reid Hoffman , Jessica Livingston , Peter Thiel , Amazon Web Services (AWS), and Infosys . However, the actual collected funds fell considerably short of these initial pledges, amounting to only $130 million by 2019, according to company disclosures.

From its inception, OpenAI articulated a mission focused on open collaboration, intending to share patents and research. This vision, however, evolved, and the organization later began to restrict access to its most advanced models, citing competitive pressures and safety considerations. The early days of OpenAI saw its operations commence from Greg Brockman’s living room, before eventually establishing its headquarters at the Pioneer Building in the Mission District, San Francisco .

The founding charter of OpenAI clearly stated its primary objective: “to ensure that artificial general intelligence (AGI)—by which we mean highly autonomous systems that outperform humans at most economically valuable work—benefits all of humanity.” This mission was partly driven by concerns about AI safety and the potential existential risk from artificial general intelligence , as articulated by Musk and Altman in 2015. They recognized both the immense potential for AI to benefit society and the equally significant risks if it were developed or deployed improperly. The organization emphasized that AI should serve as an extension of human will, distributed broadly and equitably, in the spirit of liberty. Recognizing the unpredictable nature of AI development, OpenAI aimed to be a leading research institution that would prioritize positive outcomes for all, rather than narrow self-interest. Sam Altman, a co-chair, anticipated a decades-long endeavor that would ultimately lead to AI surpassing human intelligence.

To assemble a world-class research team, Greg Brockman engaged with influential figures in the AI community, such as Yoshua Bengio , a pioneer in deep learning . This outreach resulted in the recruitment of nine leading AI researchers as OpenAI’s initial employees in December 2015. Although OpenAI could not match the salaries and stock options offered by tech giants like Facebook and Google , its compelling mission and potential attracted top talent. In 2016 alone, the company allocated $7 million for its first 52 employees. The allure of OpenAI’s vision was so strong that some researchers, like Wojciech Zaremba , reportedly turned down offers significantly exceeding their market value to join.

In April 2016, OpenAI launched “OpenAI Gym,” a public beta platform designed to facilitate reinforcement learning research. Later that year, in August, Nvidia donated its first DGX-1 supercomputer to OpenAI, significantly accelerating the training of complex AI models by reducing processing times from days to mere hours. December 2016 saw the release of “Universe,” a software platform intended to measure and train an AI’s general intelligence across a wide array of digital environments, including games and websites.

Corporate Structure and Evolution

The organizational structure of OpenAI underwent a significant transformation in 2019. It transitioned from its original non-profit status to a “capped” for-profit model, where profits were limited to 100 times the initial investment. This strategic shift, according to OpenAI, was designed to enable OpenAI Global, LLC, to attract necessary investment from venture capital firms and to grant employees equity stakes in the company. This was a crucial move, as many leading AI researchers were drawn to companies like Google Brain , DeepMind , and Facebook precisely because of the stock options they offered, which were unavailable to a traditional non-profit. Prior to this change, OpenAI was legally obligated to publicly disclose the compensation of its top employees.

Following this transition, OpenAI began distributing equity to its staff and solidified a pivotal partnership with Microsoft, announcing a $1 billion investment. This collaboration led to OpenAI’s systems being powered by Microsoft’s Azure -based supercomputing infrastructure. OpenAI Global, LLC then declared its intention to commercialize its technologies, projecting an expenditure of $1 billion within five years, a figure that Altman suggested might ultimately prove insufficient, given the immense capital required to achieve AGI.

Under this new structure, the nonprofit OpenAI, Inc. maintained its position as the sole controlling shareholder of the for-profit OpenAI Global, LLC. Despite its for-profit designation, OpenAI Global, LLC was bound by a formal fiduciary responsibility to uphold OpenAI, Inc.’s original nonprofit charter. Safeguards were put in place, including barring a majority of OpenAI, Inc.’s board from holding financial stakes in the for-profit entity, and recusing minority stakeholders with financial interests from certain voting matters due to potential conflicts. Nevertheless, some researchers voiced concerns that the move to a for-profit model was at odds with OpenAI’s stated goal of “democratizing” AI.

A significant legal challenge emerged on February 29, 2024, when Elon Musk filed a lawsuit against OpenAI and Sam Altman. Musk alleged that the company had deviated from its public benefit mission in favor of profit maximization. OpenAI dismissed the suit as “incoherent” and “frivolous,” though Musk later reignited legal action against Altman and others in August. In response, on April 9, 2024, OpenAI filed a countersuit against Musk, accusing him of employing “bad-faith tactics” to impede the company’s progress and attempting to seize its innovations for his personal gain. OpenAI also asserted that Musk had previously supported the establishment of a for-profit structure and had expressed a desire for control over the company. The countersuit sought damages and injunctive relief to prevent further alleged interference. In a further development on February 10, 2025, a consortium of investors led by Elon Musk submitted an unsolicited bid of $97.4 billion to acquire the nonprofit entity controlling OpenAI, expressing readiness to match or exceed any superior offers. This bid was rejected on February 14, 2025, with OpenAI reiterating its status as not being for sale, though the offer complicated Altman’s restructuring plans by potentially influencing the valuation benchmark for the nonprofit.

The original nonprofit design of OpenAI was rooted in the principle that AGI should “benefit all of humanity,” rather than accrue to “the private gain of any person.” In 2019, this led to the creation of OpenAI Global, LLC, a capped-profit subsidiary under the nonprofit’s control. By December 2024, OpenAI proposed a significant restructuring plan. This involved converting the capped-profit entity into a Delaware-based public benefit corporation (PBC) and releasing it from the direct control of the nonprofit. Under this proposed arrangement, the nonprofit would divest its control and other assets, receiving equity in return. This equity would then be used to fund and pursue separate charitable initiatives, particularly in science and education. OpenAI leadership characterized this change as essential for securing additional investment and argued that the nonprofit’s founding mission would ultimately be better served. However, the plan drew criticism from former employees, with a legal letter titled “Not For Private Gain” urging the attorneys general of California and Delaware to intervene. The letter contended that the restructuring was illegal, would strip away governance safeguards, and that the nonprofit was best positioned to advance its mission by retaining control of OpenAI Global, LLC, regardless of the equity received.

2025 Restructuring and Microsoft Partnership

On October 28, 2025, OpenAI announced the formal adoption of its new PBC corporate structure, having secured approval from the attorneys general of California and Delaware. This restructuring saw the for-profit arm rebranded as OpenAI Group PBC, while the nonprofit entity was renamed the OpenAI Foundation. In this new configuration, the OpenAI Foundation holds a 26% stake in the PBC, Microsoft maintains a 27% stake, and employees and other investors collectively own the remaining 47%. All board members of OpenAI Group PBC are appointed by the OpenAI Foundation, which possesses the authority to remove them at any time. Members of the Foundation’s board are also designated to serve on the for-profit board. This revised structure empowers the for-profit PBC to attract investor capital in a manner akin to traditional tech companies, including the potential for an initial public offering , a path that Altman indicated was the most probable future direction.

The partnership with Microsoft, a cornerstone of OpenAI’s operational capabilities, has been deeply intertwined with its growth. By January 2023, OpenAI was reportedly in discussions for a funding round that would value the company at $29 billion, a doubling of its 2021 valuation. On January 23, 2023, Microsoft formally announced a significant new investment of US$10 billion in OpenAI Global, LLC, to be disbursed over several years. A substantial portion of this investment was tied to OpenAI’s utilization of Microsoft’s Azure cloud computing services. Between September and December 2023, Microsoft undertook a comprehensive rebranding of its Copilot offerings, integrating them across various platforms, including Windows installations and mobile applications.

Following the 2025 restructuring, Microsoft’s stake in the for-profit OpenAI Group PBC was valued at $135 billion. Concurrently, OpenAI committed to a substantial agreement to purchase $250 billion worth of Azure services, with Microsoft relinquishing its right of first refusal on OpenAI’s future cloud computing acquisitions. This multifaceted deal also stipulated that OpenAI would continue to share 20% of its revenue with Microsoft until the achievement of AGI, which must now be validated by an independent expert panel. Furthermore, the agreement relaxed restrictions on both companies’ collaborations with third parties, enabling Microsoft to pursue AGI independently and allowing OpenAI greater flexibility in developing products with other entities.

Financial Landscape

The financial operations of OpenAI are characterized by substantial investments and ambitious growth targets. In 2017, cloud computing costs alone represented a quarter of OpenAI’s functional expenses, amounting to $7.9 million. This contrasts sharply with the $442 million in total expenses reported by DeepMind for the same year. The significant computational demands of training OpenAI’s advanced AI models were starkly illustrated in the summer of 2018, when the training of its Dota 2 bots required the rental of 128,000 CPUs and 256 GPUs from Google for an extended period.

By October 2024, OpenAI had successfully concluded a capital raise of $6.6 billion, valuing the company at $157 billion, with contributions from key investors including Microsoft and Nvidia. A significant development in early 2025 was the announcement by Donald Trump of The Stargate Project , a public-private partnership involving OpenAI, Oracle , SoftBank , and MGX . This ambitious initiative aims to construct an AI infrastructure system in collaboration with the US government , with an estimated cost of $500 billion to be funded over four years. In July 2025, the United States Department of Defense awarded OpenAI, along with Anthropic, Google, and xAI , a $200 million contract for AI development in military applications. Simultaneously, OpenAI entered into an agreement with the UK Government to deploy ChatGPT and other AI tools within public services. To support its broader mission, OpenAI also established a $50 million fund dedicated to assisting nonprofit and community organizations.

Further bolstering its financial standing, OpenAI secured $40 billion in funding in April 2025, achieving a post-money valuation of $300 billion. This transaction represented the largest private technology deal in history, with SoftBank leading the round and participation from prominent investors such as Microsoft, Coatue , Altimeter , and Thrive . By July 2025, the company reported an annualized revenue of $12 billion, a substantial increase from the $3.7 billion recorded in 2024. This surge was largely attributed to the growing success of ChatGPT subscriptions, which reached 20 million paid users by April 2025, and a rapidly expanding enterprise customer base that exceeded five million business users.

Despite this impressive revenue growth, OpenAI’s cash burn remains exceptionally high, driven by the immense computational costs associated with training and operating large language models. The company projects an operating loss of $8 billion for 2025. Looking ahead, OpenAI has revised its long-term spending projections to approximately $115 billion through 2029, with annual expenditures anticipated to escalate significantly, reaching $17 billion in 2026, $35 billion in 2027, and $45 billion in 2028. These substantial investments are earmarked for expanding compute infrastructure, developing proprietary AI chips, constructing data centers, and funding intensive model training programs. More than half of this projected spending through the end of the decade is expected to support research-intensive compute operations crucial for model development. This aggressive financial strategy reflects OpenAI’s prioritization of market expansion and technological leadership over immediate profitability, with a target of achieving cash flow positive operations by 2029 and projecting revenues of around $200 billion by 2030. In October 2025, an employee share sale valued the company at $500 billion, cementing its status as the world’s most valuable privately held company, surpassing SpaceX .

The Sam Altman Ouster and its Aftermath

A dramatic turn of events unfolded on November 17, 2023, when Sam Altman was abruptly removed as CEO by the OpenAI board of directors, citing a “lack of confidence.” The board at the time consisted of Helen Toner , Ilya Sutskever , Adam D’Angelo , and Tasha McCauley. Chief Technology Officer Mira Murati assumed the role of interim CEO. Greg Brockman , the president of OpenAI, was also removed as chairman of the board and subsequently resigned from his presidential position. This upheaval triggered the resignation of three senior OpenAI researchers: Jakub Pachocki, the director of research and lead on GPT-4; Aleksander Mądry [pl], the head of AI risk; and researcher Szymon Sidor.

The following day, November 18, 2023, investor pressure, notably from Microsoft and Thrive Capital , began to mount for Altman’s reinstatement. Reports emerged of talks regarding his return, with Altman himself indicating a willingness to consider starting a new company with former OpenAI employees if negotiations failed. The board members reportedly agreed “in principle” to resign should Altman be reinstated. However, by November 19, 2023, negotiations had stalled, and Emmett Shear was appointed as the new interim CEO. In the interim, the board had also approached Anthropic CEO Dario Amodei , a former OpenAI executive, about potentially replacing Altman and even exploring a merger between the two companies, but both offers were declined.

On November 20, 2023, Microsoft CEO Satya Nadella announced that Altman and Brockman would join Microsoft to lead a new advanced AI research team, while also expressing continued commitment to OpenAI. Before this partnership was finalized, Altman reportedly offered the board another opportunity to negotiate. This move coincided with a near-total employee revolt, as approximately 738 of OpenAI’s 770 employees, including Murati and Sutskever, signed an open letter threatening to resign and join Microsoft if Altman was not rehired and the board did not step down. This action prompted OpenAI investors to explore legal recourse against the board. Amidst the escalating crisis, OpenAI management sent an internal memo indicating that negotiations with Altman and the board had resumed.

The intense standoff concluded on November 21, 2023, with Altman and Brockman returning to their respective roles at OpenAI. This resolution was accompanied by a significantly reconstructed board, including new members Bret Taylor (as chairman) and Lawrence Summers , with Adam D’Angelo remaining. Concerns regarding Altman’s handling of potential safety implications of the company’s discoveries were reportedly raised with the board prior to his initial firing. On November 29, 2023, OpenAI announced the addition of an anonymous Microsoft employee to the board as a non-voting observer, though Microsoft later resigned from this role in July 2024. In February 2024, the Securities and Exchange Commission issued a subpoena for OpenAI’s internal communications to investigate whether Altman’s alleged lack of candor had misled investors. The period following Altman’s temporary ouster and subsequent return saw a gradual departure of employees, including many key members of the original leadership and a substantial number of AI safety researchers.

Acquisitions and Corporate Partnerships

OpenAI has strategically expanded its capabilities and market reach through a series of acquisitions. In August 2023, the company acquired Global Illumination, a New York-based startup focused on developing digital infrastructure and creative tools using AI. This was followed in June 2024 by the acquisition of Multi, a startup specializing in remote collaboration technologies. In March 2025, OpenAI entered into an agreement with CoreWeave , acquiring $350 million worth of CoreWeave shares and gaining access to AI infrastructure in exchange for $11.9 billion over five years. Notably, Microsoft was already CoreWeave’s largest customer in 2024. Simultaneously, OpenAI and Microsoft were renegotiating the terms of their partnership to facilitate OpenAI’s potential future initial public offering while ensuring Microsoft’s continued access to advanced AI models.

A significant move occurred on May 21, when OpenAI announced the $6.5 billion acquisition of io , an AI hardware startup founded in 2024 by former Apple designer Jony Ive . In September 2025, OpenAI agreed to acquire the product testing startup Statsig for $1.1 billion in an all-stock deal, appointing Statsig’s CEO, Vijaye Raji , as OpenAI’s chief technology officer of applications. The company also revealed its development of an AI-driven hiring service, poised to challenge LinkedIn . OpenAI further expanded its consumer-facing offerings by acquiring the personal finance app Roi in October 2025. In December 2025, it was announced that OpenAI had agreed to acquire Neptune, an AI tooling startup that assists companies in tracking and managing model training, for an undisclosed sum.

OpenAI has also cultivated a wide array of corporate partnerships. The company faced criticism for outsourcing the annotation of data sets to Sama, a San Francisco-based firm employing workers in Kenya . These annotations were crucial for training AI models designed to detect toxicity, particularly for moderating content from ChatGPT. However, the annotated texts often contained graphic descriptions of violence, including sexual violence. Investigations revealed that OpenAI began providing data to Sama as early as November 2021, with Sama employees reporting significant psychological distress. While OpenAI paid Sama $12.50 per hour, the annotators received the equivalent of $1.32 to $2.00 per hour after taxes, with Sama citing infrastructure and quality assurance costs contributing to the higher rate.

In 2024, OpenAI initiated a collaboration with Broadcom to design a custom AI chip optimized for both training and inference, with mass production targeted for 2026 by TSMC using a 3 nm node. This initiative aims to reduce OpenAI’s reliance on Nvidia GPUs, which are subject to high demand and cost. In January 2024, Arizona State University became the first university to adopt ChatGPT Enterprise, marking OpenAI’s initial foray into academic partnerships. A landmark agreement was reached in June 2024 when Apple Inc. partnered with OpenAI to integrate ChatGPT features into its products as part of its new Apple Intelligence initiative. In June 2025, OpenAI began utilizing Google Cloud ’s Tensor Processing Units (TPUs ) to support ChatGPT and related services, signaling a diversification of its hardware reliance beyond Nvidia. By September 2025, OpenAI had signed a substantial contract with Oracle, committing to purchase $300 billion in computing power over the next five years. In the same month, OpenAI and NVIDIA announced a significant partnership, including the potential deployment of at least 10 gigawatts of NVIDIA systems and a $100 billion investment from NVIDIA into OpenAI. A multi-billion dollar deal with AMD was announced in October 2025, wherein OpenAI committed to purchasing six gigawatts of AMD chips and secured an option to acquire up to 10% of AMD’s shares. In December 2025, Disney announced a $1 billion investment in OpenAI and a three-year licensing agreement allowing users to generate videos using Sora, with access to over 200 Disney, Marvel , Star Wars , and Pixar characters.

Government Contracting and Services

OpenAI actively engages with government entities, providing its LLMs to programs such as the Artificial Intelligence Cyber Challenge and the Advanced Research Projects Agency for Health . In October 2024, The Intercept revealed that OpenAI’s tools were deemed “essential” for AFRICOM ’s mission, integrated into a contractual agreement between the United States Department of Defense and Microsoft . By December 2024, OpenAI announced a partnership with defense-tech company Anduril to develop drone defense technologies for the United States and its allies. In 2025, Kevin Weil, OpenAI’s Chief Product Officer, was commissioned as a lieutenant colonel in the U.S. Army to serve as a senior advisor for Detachment 201 .

In June 2025, the U.S. Department of Defense awarded OpenAI a $200 million one-year contract for the development of AI tools tailored for military and national security applications. In response, OpenAI launched “OpenAI for Government,” a program designed to grant federal, state, and local governments access to its models, including ChatGPT.

Products and Development Milestones

OpenAI’s product portfolio has expanded significantly, marked by key developments in language, image, and video generation. The release of GPT-2 in February 2019 garnered considerable attention for its remarkable ability to produce human-like text. In 2020, OpenAI introduced GPT-3 , a language model trained on vast internet datasets, designed for natural language question answering, translation, and coherent text generation. The associated API, also named the API, was launched as OpenAI’s inaugural commercial product.

A notable development in 2021 was the introduction of DALL-E , a specialized deep learning model capable of generating complex digital images from textual descriptions, utilizing a derivative of the GPT-3 architecture. The subsequent release of ChatGPT in November 2022 proved to be a watershed moment in the AI boom , rapidly becoming the fastest-growing consumer application in history, attracting over 100 million users within its first two months.

The widespread media coverage following ChatGPT’s free preview launch in December 2022, which saw over a million sign-ups in five days, spurred competitors into action. Google, concerned about ChatGPT’s potential to disrupt its dominance as an information source, announced its own AI application, Bard . In February 2023, Microsoft revealed its strategy to integrate AI technology based on the same foundation as ChatGPT into its suite of products, including Microsoft Bing , Edge , and Microsoft 365 .

March 14, 2023, marked the release of GPT-4 , made available through an API with a waitlist and as a feature within ChatGPT Plus. Building on this, November 6, 2023, saw the launch of GPTs, enabling users to create customized versions of ChatGPT for specific applications, thereby broadening the scope of AI’s industrial utility. The overwhelming demand for ChatGPT Plus led to a temporary suspension of new sign-ups on November 14, 2023, with access for new subscribers eventually resuming on December 13.

December 2024 was a pivotal month, featuring the launch of the Sora model, a text-to-video generation system. OpenAI also introduced OpenAI o1 , an early reasoning model internally codenamed “strawberry,” and unveiled ChatGPT Pro, a premium subscription service offering unlimited o1 access and enhanced voice capabilities. Preliminary benchmark results for the forthcoming OpenAI o3 models were also shared.

In January 2025, OpenAI released Operator, an AI agent designed for web automation, allowing users to define goals for the tool to execute on websites. Initially available only to Pro users in the United States, this marked a significant step towards more autonomous AI interaction. Nine days later, the deep research agent was launched, achieving a 27% accuracy rate on the Humanity’s Last Exam benchmark. Sam Altman later indicated that GPT-4.5 would be the final model to lack full chain-of-thought reasoning. By July 2025, reports suggested that AI models from both OpenAI and Google DeepMind had demonstrated the capability to solve mathematics problems at the level of top performers in the International Mathematical Olympiad , with OpenAI’s model achieving gold medal-equivalent performance.

During the company’s DevDay event on October 6, 2025, OpenAI unveiled its Agent Builder platform, featuring a drag-and-drop interface to facilitate the design, testing, and deployment of agentic workflows for developers and businesses, irrespective of extensive coding expertise. On October 21, 2025, OpenAI introduced ChatGPT Atlas , a browser designed to integrate the ChatGPT assistant directly into web navigation, aiming to compete with established browsers like Google Chrome and Apple Safari . The year concluded on December 11, 2025, with the announcement of GPT-5.2, a model slated for enhanced capabilities in spreadsheet creation, presentation building, image perception, code writing, and long-context understanding.

Transparency, Alignment, and Data Practices

OpenAI has faced scrutiny regarding its transparency, particularly concerning the technical details of its products. In March 2023, criticisms arose due to the limited disclosure of information about models like GPT-4, which some argued contradicted the company’s initial commitment to openness and hindered independent research and safeguard development. OpenAI cited competitive and safety concerns as justifications for this shift. Former chief scientist Ilya Sutskever posited in 2023 that the open-sourcing of increasingly powerful models posed growing risks, suggesting that the rationale for withholding the most potent AI models would become evident over time.

In September 2025, OpenAI published a study analyzing user interactions with ChatGPT for everyday tasks. The study indicated that “non-work tasks,” as classified by an LLM, constituted over 72 percent of ChatGPT usage, with a smaller proportion related to business productivity.

The organization’s commitment to AI alignment was highlighted in July 2023 with the launch of the superalignment project. This initiative aimed to develop methods for aligning future superintelligences within four years, by automating alignment research through AI itself. OpenAI pledged 20% of its computing resources to this project, though team members denied receiving a commensurate allocation. The superalignment project was ultimately terminated in May 2024, following the departure of its co-leaders, Ilya Sutskever and Jan Leike .

In August 2025, OpenAI faced criticism due to the inadvertent exposure of thousands of private ChatGPT conversations to public search engines. An experimental feature, intended to allow users to make specific chats discoverable, resulted in some personal details appearing in search results. OpenAI promptly removed the feature and began working with search providers to remove the exposed content, characterizing it as a design flaw rather than a security breach. CEO Sam Altman acknowledged the issue, noting the sensitive nature of user interactions with ChatGPT.

Management and Personnel

The leadership structure at OpenAI is extensive and dynamic. Sam Altman , a co-founder and former president of Y Combinator , serves as CEO. Greg Brockman , another co-founder and former CTO, holds the position of President. Key figures in research and development include Chief Scientist Officer Jakub Pachocki , formerly Director of Research, and Chief Research Officer Mark Chen. Operational leadership is managed by Chief Operating Officer Brad Lightcap, with Sarah Friar , former CEO of Nextdoor , serving as CFO. Kevin Weil, previously at Twitter, Inc. and Meta Platforms , is the Chief Product Officer. Scott Schools , formerly of Uber , holds the role of Chief Compliance Officer, while Chris Lehane , formerly of Airbnb , leads global affairs. Aaron Chatterji , a professor at Duke University’s Fuqua School of Business , serves as Chief Economist. Fidji Simo , former CEO of Instacart , leads the Applications division.

The board of directors for the OpenAI nonprofit includes Bret Taylor as chairman, alongside Sam Altman , Adam D’Angelo , Sue Desmond-Hellmann , Nicole Seligman , Paul Nakasone , Zico Kolter , and Adebayo Ogunlesi .

Prominent individual investors associated with OpenAI include Reid Hoffman (LinkedIn co-founder), Peter Thiel (PayPal co-founder), Jessica Livingston (Y Combinator founding partner), and co-founder Elon Musk .

The company has experienced significant personnel changes. In 2018, Elon Musk resigned from the board due to potential conflicts of interest with his role at Tesla and its AI development for self-driving cars . Reid Hoffman resigned from the board in March 2023 to avoid conflicts with his investments in other AI companies. In May 2024, Chief Scientist Ilya Sutskever resigned, succeeded by Jakub Pachocki , with co-leader Jan Leike also departing amid trust concerns. Paul Nakasone joined the board in the same period. August 2024 saw co-founder John Schulman leave for Anthropic , while president Greg Brockman took an extended leave. In September 2024, CTO Mira Murati departed the company. Notably, Lawrence Summers resigned from the board in November 2025.

Governance and Legal Issues

In May 2023, Altman, Brockman, and Sutskever published recommendations for the governance of superintelligence , predicting its potential emergence within a decade and emphasizing the need for proactive measures against existential risks. They proposed an international watchdog organization, akin to the IAEA , to oversee advanced AI systems, alongside increased technical safety research and international coordination.

In July 2023, the FTC initiated a civil investigative demand to investigate OpenAI’s data security and privacy practices concerning ChatGPT, examining potential violations of Section 5 of the Federal Trade Commission Act of 1914 , including allegations of unfair practices and harm to consumers through reputational damage. The FTC also expressed concern regarding circular spending arrangements between Microsoft and OpenAI, particularly the exchange of Azure credits for engineering talent, citing potential negative public impacts.

OpenAI’s global affairs chief endorsed the UK’s approach to AI regulation in September 2024, testifying before a House of Lords committee. In February 2025, Sam Altman expressed interest in collaborating with the People’s Republic of China, despite U.S. regulatory restrictions, a move seen as a response to the growing influence of Chinese AI companies like DeepSeek . OpenAI subsequently enhanced its security protocols to mitigate risks of industrial espionage .

Advocacy for federal preemption of state AI laws has been a notable stance for OpenAI, with the company opposing California’s AI legislation and arguing for a stronger federal role. This position has been met with opposition from organizations like Public Citizen , which argued that existing state laws have not hindered innovation.

Non-Disparagement Agreements and Copyright Disputes

Prior to May 2025, OpenAI required departing employees to sign lifelong non-disparagement agreements that prohibited criticism of the company and acknowledged the agreement’s existence. Former employee Daniel Kokotajlo publicly stated he forfeited vested equity to avoid signing. Sam Altman claimed ignorance of the equity cancellation provision, though leaked documents reportedly contradicted this. OpenAI later issued a memo releasing former employees from these agreements.

OpenAI has faced significant legal challenges regarding copyright infringement . In July 2023, authors including Sarah Silverman, Matthew Butterick , Paul Tremblay, and Mona Awad filed lawsuits. In September 2023, a class-action lawsuit was filed by 17 authors, including George R. R. Martin and John Grisham , alleging illegal use of copyrighted works for training AI. The New York Times also initiated a lawsuit in late December 2023. In May 2024, it was revealed that OpenAI had destroyed its Books1 and Books2 training datasets, believed by the Authors Guild to contain over 100,000 copyrighted books.

Concerns were also raised internally regarding the use of YouTube videos for training GPT-4 via the Whisper speech recognition tool, potentially violating YouTube’s terms of service. In February 2024, The Intercept, Raw Story, and Alternate Media Inc. filed copyright litigation against OpenAI. On April 30, 2024, eight newspapers, including The Mercury News and the Chicago Tribune , sued OpenAI and Microsoft for allegedly harvesting copyrighted articles. A June 2023 lawsuit claimed OpenAI scraped 300 billion words online without consent.

In May 2024, OpenAI reached a content integration agreement with News Corp , while other publications opted for legal action. A coalition of Canadian news outlets filed a similar lawsuit in November 2024. In October 2024, Suchir Balaji , a former OpenAI employee involved in copyright litigation, died by suicide, prompting calls for an investigation into potential foul play.

GDPR Compliance and Military Use

In April 2023, the European Data Protection Board established a task force to address cross-border data transfer issues and potential enforcement actions related to ChatGPT, following actions by the Italian data protection authority. In late April 2024, NOYB filed a complaint concerning alleged violations of the European General Data Protection Regulation by OpenAI, citing instances of ChatGPT generating false personal information and denying correction requests.

OpenAI faced criticism for lifting its ban on using ChatGPT for “military and warfare” activities in January 2024. The revised usage policies prohibit activities with a high risk of physical harm, including weapons development and warfare, though the specific wording shifted from a direct ban to a prohibition on causing harm and developing weapons.

Wrongful Death Lawsuits (2025)

In August 2025, a wrongful death lawsuit was filed against OpenAI and Sam Altman by the parents of a teenager who died by suicide, alleging that prolonged conversations with ChatGPT about mental health and self-harm contributed to his death. OpenAI expressed condolences and committed to enhancing safety measures. This case was described as the first of its kind targeting a chatbot company. In November 2025, the Social Media Victims Law Center filed seven lawsuits, including four wrongful death claims, on behalf of individuals who allegedly committed suicide after extensive ChatGPT usage.

See Also

• Anthropic – A direct competitor in the AI research space.
• Google DeepMind – A major AI research laboratory and competitor.
• xAI – Another significant artificial intelligence corporation.