Jump to content

OpenAI

From Alfawiki
OpenAI
Company typePublic benefit corporation
IndustryArtificial intelligence
FoundedDecember 2015; 10 years ago (2015-12)
FounderSam Altman, Elon Musk, Ilya Sutskever, Greg Brockman
Headquarters
San Francisco
,
United States
Key people
Sam Altman (CEO)
ProductsGPT series, DALL-E, ChatGPT, Sora (text-to-video model)
OwnerOpenAI Foundation (26%), Microsoft (27%), employees and investors (47%)
Number of employees
52 (initial group in 2015)
ParentOpenAI Foundation (non-profit arm)

OpenAI is a United States-based organization specializing in artificial intelligence (AI), headquartered in San Francisco, California. Its primary objective is the development of "safe and beneficial" artificial general intelligence (AGI), which it defines as "highly autonomous systems that surpass human performance in most economically significant tasks." As a prominent entity in the ongoing AI boom, OpenAI is recognized for its GPT series of large language models, alongside the DALL-E models that generate images from text prompts, and the text-to-video model known as Sora. The introduction of ChatGPT in November 2022 notably amplified public and industry interest in generative AI technologies.

OpenAI’s corporate framework is notably intricate. By October 2025, the non-profit OpenAI Foundation—established in 2015 and incorporated in Delaware—held a 26% ownership stake in OpenAI Group PBC, a for-profit public benefit corporation responsible for commercializing OpenAI’s innovations. Microsoft has contributed over $13 billion in investments and supplies cloud computing services through Azure. In October 2025, OpenAI finalized a $6.6 billion equity sale that placed the company’s valuation at $500 billion. Subsequent to this transaction, Microsoft acquired a 27% share, while employees and investors retained 47%, and the OpenAI Foundation maintained its 26% holding.

Between 2023 and 2024, OpenAI encountered several lawsuits alleging copyright violations related to its training datasets. In November 2023, the board temporarily dismissed CEO Sam Altman but reinstated him five days later following changes to the board’s composition. Throughout 2024, approximately half of OpenAI’s researchers specializing in AI safety resigned, expressing apprehensions about the company’s influence on broader industry challenges.

Founding

Established in December 2015 as a non-profit entity, OpenAI was co-founded by Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman, among others, with Altman and Musk serving as co-chairs. The initial funding commitments amounted to $1 billion from individuals including Altman, Brockman, Musk, Reid Hoffman, Jessica Livingston, Peter Thiel, as well as companies like Amazon Web Services (AWS) and Infosys. However, by 2019 only $130 million had been received. The organization pledged to engage in "open collaboration," sharing certain patents and research openly. Its early operations took place in Brockman’s living room before relocating to the Pioneer Building located in San Francisco’s Mission District.

The founding vision centered on ensuring that artificial general intelligence (AGI)—defined as highly autonomous systems capable of outperforming humans across most economically valuable tasks—would benefit all humanity. Musk and Altman highlighted concerns regarding AI safety and potential existential threats posed by AGI. They viewed AI as an extension of human intent that should be widely accessible and emphasized establishing a research institution dedicated to positive societal impacts. Altman foresaw the endeavor spanning decades to achieve intelligence beyond human capabilities.

Despite offering salaries below those at tech giants like Google or Facebook and not providing stock options, Brockman successfully recruited nine top-tier AI researchers by December 2015. In 2016 alone, OpenAI invested $7 million on its initial group of 52 employees, attracting professionals motivated more by mission than financial incentives. Co-founder Wojciech Zaremba notably turned down offers two to three times higher elsewhere to be part of the team.

In April 2016, OpenAI introduced "OpenAI Gym," a platform designed to support reinforcement learning research. Later that year, in August, Nvidia contributed its DGX-1 supercomputer to accelerate model training processes, reducing computation time from six days down to two hours. By December 2016, OpenAI released "Universe," a software platform aimed at evaluating and training AI general intelligence across diverse environments including games, websites, and applications.

Shift from nonprofit to capped-profit model

In 2019, OpenAI transitioned from a nonprofit organization to a "capped" for-profit entity, limiting investor returns to 100 times their initial capital. This arrangement allows OpenAI Global, LLC to attract venture funding and provide equity incentives to employees, aligning its practices with industry leaders such as Google Brain and DeepMind. Previously, the organization was obligated to disclose the highest employee compensation.

After adopting this new structure, OpenAI allocated equity shares to its staff and secured a $1 billion investment from Microsoft. This partnership also involves operating OpenAI’s systems on Microsoft Azure-powered supercomputers. OpenAI Global, LLC announced intentions to commercially license its innovations and anticipated spending $1 billion over five years, with CEO Sam Altman acknowledging that even this investment might fall short of achieving artificial general intelligence (AGI).

The nonprofit parent entity, OpenAI, Inc., maintains majority control over OpenAI Global, LLC and enforces fiduciary responsibilities consistent with its nonprofit mission. Most board members are barred from holding financial interests in the for-profit subsidiary, and conflict-of-interest policies restrict certain voting privileges. Nonetheless, some industry analysts question whether the for-profit structure is fully compatible with OpenAI’s goal of democratizing artificial intelligence.

Legal conflicts have emerged involving co-founder Elon Musk. In February 2024, Musk filed a lawsuit against OpenAI and CEO Sam Altman, alleging a shift away from public benefit toward profit-driven motives; OpenAI dismissed the case as baseless. Musk renewed legal actions in August 2024. In response, OpenAI filed a countersuit in April 2024 accusing Musk of bad-faith conduct intended to obstruct the company’s progress. In February 2025, an investor group led by Musk proposed a $97.4 billion acquisition of the nonprofit controlling OpenAI; the offer was declined, complicating plans for organizational restructuring.

Originally established as a nonprofit to ensure that AGI benefits all humanity rather than private entities, OpenAI created its capped-profit subsidiary in 2019. In December 2024, it announced plans to restructure by converting the capped-profit entity into a Delaware public benefit corporation (PBC), thereby releasing it from nonprofit oversight. The nonprofit would divest its ownership stake in exchange for equity to support separate charitable programs focused on science and education. Leadership contends that this change is necessary to attract additional investment while preserving the original mission.

This proposed restructuring has drawn criticism from former employees and advocacy organizations who argue it weakens governance protections and undermines the nonprofit’s founding purpose. A legal petition urged state attorneys general to intervene, asserting that the current complex structure was designed to maintain accountability without profit conflicts. Critics warn that PBCs may prioritize financial returns over mission depending on the influence of controlling shareholders.

Legal authorities emphasize that altering OpenAI’s stated purpose would require proof that its existing mission has become illegal or impractical to fulfill.

2025 corporate reorganization

On October 28, 2025, OpenAI revealed a revised corporate framework that received approval from the attorneys general of California and Delaware. The organization’s for-profit division transformed into a public benefit corporation named OpenAI Group PBC, while its non-profit arm was rebranded as the OpenAI Foundation. The Foundation retains a 26% ownership in the PBC, Microsoft holds 27%, and the remaining 47% belongs to employees and other investors. The Foundation holds authority to appoint and remove all board members of the PBC, with some directors serving on both boards. This arrangement permits OpenAI Group PBC to attract capital from investors similarly to conventional technology companies, including through an initial public offering—a path CEO Sam Altman indicated as probable.

Collaboration with Microsoft

In January 2023, OpenAI Global, LLC secured funding that valued the company at $29 billion, effectively doubling its worth from 2021. During that same month, Microsoft declared a $10 billion long-term investment in OpenAI Global, LLC, partly intended to enhance usage of its Azure cloud services.

By September 2023, Microsoft consolidated its artificial intelligence products under the Microsoft Copilot brand, merging Bing Chat with Microsoft 365 Copilot. This initiative expanded in December 2023 when Microsoft integrated Copilot into Windows 10 and 11 systems and launched standalone applications for Android and iOS platforms.

Following the restructuring in 2025, Microsoft acquired a 27% equity stake in OpenAI Group PBC, which was then valued at $135 billion. Simultaneously, OpenAI committed to purchasing $250 billion worth of Azure services from Microsoft; in return, Microsoft waived its right of first refusal on future cloud contracts. Under this agreement, OpenAI will share 20% of its revenue with Microsoft until achieving artificial general intelligence (AGI), which is now subject to independent validation. Restrictions on joint efforts were relaxed, enabling both entities to independently pursue AGI development and collaborate with other partners on product innovation.

Financial Overview

In 2017, OpenAI allocated $7.9 million—representing one-quarter of its operational expenses—to cloud computing costs, contrasting with DeepMind’s total expenditures of $442 million that year. Training OpenAI’s Dota 2 bots in 2018 involved leasing 128,000 CPUs and 256 GPUs from Google over several weeks.

By October 2024, OpenAI raised $6.6 billion at a valuation of $157 billion through investments from Microsoft, Nvidia, and SoftBank.

On January 21, 2025, former President Donald Trump announced The Stargate Project, a $500 billion AI infrastructure collaboration involving OpenAI, Oracle, SoftBank, MGX, and the U.S. government to be financed over four years. In July of that year, OpenAI secured a $200 million contract with the Department of Defense alongside Anthropic, Google, and xAI. That same month marked the launch of a strategic alliance with the UK government aimed at deploying ChatGPT and AI technologies within public sector services. Additionally, OpenAI introduced a $50 million fund dedicated to supporting nonprofits and community initiatives.

In April 2025, OpenAI closed a record-breaking $40 billion funding round led by SoftBank at a post-money valuation of $300 billion—the largest private technology financing ever—joined by investors including Microsoft, Coatue Management, Altimeter Capital, and Thrive Capital.

By July 2025, annualized revenues reached $12 billion, up significantly from $3.7 billion in the previous year. This surge was driven by ChatGPT’s subscription base expanding to 20 million paid users by April and an enterprise clientele growing to five million business customers.

Despite robust revenue increases, OpenAI continues to experience substantial cash outflows due to the high computational demands associated with training large language models. The company anticipates an $8 billion operating loss for 2025. Projections indicate total expenditures will climb to around $115 billion through 2029—$80 billion more than earlier forecasts—with yearly cash burn expected to rise progressively: $17 billion in 2026, $35 billion in 2027, and reaching $45 billion in 2028. These funds are primarily allocated toward scaling compute infrastructure, developing proprietary AI hardware, constructing data centers, and supporting model training initiatives; over half of this budget is dedicated to research-focused computing resources.

OpenAI’s financial approach emphasizes expanding market presence and technological innovation over short-term profitability goals. The company aims to achieve positive cash flow by 2029 while forecasting revenues approaching $200 billion by 2030. This aggressive investment strategy highlights the substantial capital requirements involved in advancing state-of-the-art AI technology and underscores OpenAI’s commitment to sustaining industry leadership.

In October 2025, OpenAI completed an employee share sale that valued the company at $500 billion—the highest valuation ever recorded for a private firm worldwide—surpassing even SpaceX.

Termination of Altman

On November 17, 2023, the board of directors at OpenAI—comprising Helen Toner, Ilya Sutskever, Adam D'Angelo, and Tasha McCauley—dismissed Sam Altman from his role as CEO, citing a loss of confidence. Following this decision, CTO Mira Murati was appointed as interim CEO. Greg Brockman, who served as president and co-founder, was removed from his position as chairman and subsequently resigned. Additionally, three senior researchers resigned: research director Jakub Pachocki, AI risk lead Aleksander Mądry, and researcher Szymon Sidor.

Investor influence from Microsoft and Thrive Capital on November 18 led to calls for Altman’s reinstatement. Altman expressed readiness to return but also considered founding a new company with former OpenAI employees if talks failed. The board indicated they would step down "in principle" should Altman be reinstated. However, negotiations broke down on November 19, resulting in Emmett Shear replacing Murati as interim CEO. The board also approached Anthropic CEO Dario Amodei regarding a potential replacement for Altman and discussed a merger, both of which were declined.

On November 20, Microsoft CEO Satya Nadella revealed that Altman and Brockman would spearhead a new advanced AI research division within Microsoft while maintaining ties with OpenAI. Approximately 738 out of 770 OpenAI staff members, including Murati and Sutskever, signed a letter threatening to resign and join Microsoft unless the board rehired Altman and resigned themselves. This development prompted investors to contemplate legal measures against the board. OpenAI’s management confirmed ongoing discussions involving Altman and the board.

Following persistent negotiations on November 21, Altman was reinstated as CEO and Brockman resumed his role as president. The board was restructured to include new members Bret Taylor (chairman) and Lawrence Summers alongside D’Angelo. Prior to his dismissal, concerns about Altman's management of AI safety advancements had reportedly been raised with the board. On November 29, an anonymous Microsoft employee joined OpenAI’s board as a non-voting observer; however, Microsoft withdrew from this position in July 2024.

In February 2024, the Securities and Exchange Commission issued a subpoena for OpenAI’s internal communications to investigate whether Altman provided misleading information to investors regarding his transparency.

The aftermath in 2024 saw considerable turnover at OpenAI, including departures among much of the founding leadership and numerous AI safety researchers.

Acquisitions

In August 2023, OpenAI acquired Global Illumination, a New York-based startup specializing in AI-powered digital infrastructure and creative technologies.

In June 2024, OpenAI completed the acquisition of Multi, a company focused on remote collaboration solutions.

By March 2025, OpenAI secured an agreement with CoreWeave involving $350 million in equity and access to AI infrastructure, committing $11.9 billion over five years. During this period, Microsoft—CoreWeave’s largest client in 2024—was renegotiating partnership terms with OpenAI to facilitate a potential IPO while retaining access to advanced AI models.

On May 21, 2025, OpenAI announced its $6.5 billion purchase of io, an AI hardware startup founded by former Apple designer Jony Ive.

In September 2025, OpenAI agreed to acquire Statsig, a product testing startup, for $1.1 billion in an all-stock transaction. Following the acquisition, Statsig’s CEO Vijaye Raji was appointed OpenAI’s chief technology officer for applications. The company also unveiled plans for an AI-driven recruitment platform aimed at competing with LinkedIn.

October 2025 saw OpenAI expand its personalized consumer AI offerings through the acquisition of Roi, a personal finance application.

Corporate collaborations

OpenAI encountered criticism for contracting data annotation tasks to Sama, a firm based in San Francisco that employs workers located in Kenya. These annotations were utilized to train an AI model focused on detecting toxic content, which plays a role in moderating outputs, including those from ChatGPT. The datasets often contained explicit violent content. OpenAI started transferring data to Sama in November 2021. Workers at Sama reported experiencing psychological strain, earning between $1.32 and $2.00 per hour after taxes, while OpenAI compensated Sama at a rate of $12.50 per hour, covering infrastructure, quality control, and management expenses.

In 2024, OpenAI formed a partnership with Broadcom to develop a specialized AI chip intended for both training and inference purposes. The plan aims for large-scale production by TSMC in 2026 using a 3 nm fabrication process. This initiative seeks to lessen dependence on expensive Nvidia GPUs amid strong market demand.

January 2024 marked OpenAI’s initial academic agreement when Arizona State University acquired ChatGPT Enterprise. Later in June, Apple incorporated ChatGPT functionalities into its product lineup as part of its Apple Intelligence program.

Additionally, in June 2025, OpenAI began leasing Google Cloud’s Tensor Processing Units (TPUs), representing its first major deployment of AI hardware beyond Nvidia chips.

September 2025 saw OpenAI secure a five-year contract valued at $300 billion with Oracle for computing power. In the same month, NVIDIA revealed a collaboration involving the deployment of at least 10 gigawatts of systems alongside a potential investment reaching $100 billion in OpenAI.

In October 2025, OpenAI finalized a multi-billion dollar agreement with AMD to acquire six gigawatts of AMD chips starting with the MI450 model. The deal also granted OpenAI an option to purchase up to 160 million AMD shares—approximately 10% ownership—contingent on performance metrics and stock price milestones.

Government contracts

OpenAI provides large language models (LLMs) to initiatives such as the Artificial Intelligence Cyber Challenge and the Advanced Research Projects Agency for Health. In October 2024, it emerged that OpenAI’s technology was considered "essential" for AFRICOM’s operations and was included under an "Exception to Fair Opportunity" contract between the U.S. Department of Defense and Microsoft. By December 2024, OpenAI had partnered with defense technology company Anduril Industries to develop drone defense systems for use by the U.S. and allied nations.

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 within Detachment 201.

June 2025 brought a $200 million one-year contract awarded by the U.S. Department of Defense to OpenAI for creating AI tools tailored to military and national security objectives. Concurrently, OpenAI introduced OpenAI for Government, enabling federal, state, and local agencies to access its models including ChatGPT.

Products

Development

The announcement of GPT-2 occurred in February 2019 and attracted attention for its ability to generate text resembling human writing.

In 2020, OpenAI introduced GPT-3, a language model trained on vast internet datasets designed for natural language processing tasks such as translation and text generation. The API associated with GPT-3 became the foundation of OpenAI’s initial commercial offerings.

Between December 2020 and January 2021, eleven employees departed from OpenAI to establish Anthropic.

In 2021, OpenAI launched DALL-E, a deep learning model capable of producing digital images from textual descriptions based on GPT-3 architecture.

December 2022 saw the release of a free preview of ChatGPT, an AI chatbot powered by GPT-3.5, which attracted over one million users within five days. Internal forecasts projected revenues of $200 million for 2023 and $1 billion for 2024.

Following ChatGPT’s success, Google announced Bard, its competing AI chatbot, driven by concerns about market competition. On February 7, 2023, Microsoft integrated ChatGPT-powered AI into Bing search engine, Edge browser, Microsoft 365 suite, and other products.

On March 14, 2023, OpenAI launched GPT-4 as an API accessible via waitlist and as part of ChatGPT Plus subscription service.

On November 6, 2023, OpenAI introduced GPTs—customizable versions of ChatGPT tailored for specific business uses—broadening AI adoption across industries. Due to overwhelming demand, new registrations for ChatGPT Plus were temporarily paused on November 14 but resumed on December 13.

In December 2024, OpenAI unveiled the Sora text-to-video system along with OpenAI o1, an early-stage reasoning model codenamed strawberry. It also launched ChatGPT Pro—a premium subscription priced at $200 per month offering unlimited o1 access and enhanced voice capabilities—and previewed forthcoming OpenAI o3 models.

On January 23, 2025, OpenAI debuted Operator, an AI agent equipped with web automation features available to Pro subscribers in the United States. This was followed by the release of ChatGPT Deep Research agent demonstrating significant benchmark achievements. CEO Sam Altman noted that GPT-4.5 would be the final model lacking comprehensive chain-of-thought reasoning.

By July 2025, OpenAI’s models attained gold medal-level performance on International Mathematical Olympiad challenges, showcasing advanced AI reasoning abilities.

In September 2025, research published by OpenAI indicated that over 72% of ChatGPT usage pertained to non-work-related activities while business productivity constituted only a minor portion.

On October 6 during DevDay event, OpenAI launched Agent Builder—a drag-and-drop interface allowing enterprises to create and deploy AI workflows without extensive programming knowledge.

On October 21, 2025, OpenAI released ChatGPT Atlas, an AI-enhanced web browser designed to rival Google Chrome and Apple Safari by embedding ChatGPT directly into web navigation functions.

Transparency

In March 2023, OpenAI encountered backlash for disclosing only limited technical information about products such as GPT-4, marking a departure from its original commitment to openness and impeding independent research and the creation of safety measures. The organization attributed this change to concerns over competition and security. Former chief scientist Ilya Sutskever remarked that releasing advanced models as open source has become increasingly hazardous, forecasting that the rationale for withholding the most advanced AI systems on safety grounds would become clear within a few years.

Alignment

OpenAI initiated the superalignment project in July 2023 with the goal of automating alignment research through AI, targeting the alignment of future superintelligent systems within a four-year timeframe. The company announced it would allocate 20% of its computational capacity to this effort; however, insiders indicated that this pledge was not fully honored. The initiative concluded in May 2024 following the exit of its co-leaders, Ilya Sutskever and Jan Leike.

Leaked conversations

In August 2025, OpenAI faced criticism after thousands of private ChatGPT dialogues were unintentionally indexed by search engines due to an experimental feature named "share with search engines." This opt-in option inadvertently allowed some conversations containing personal information to appear in search results when users mistakenly enabled it while sharing links. OpenAI permanently disabled this feature on August 1, 2025, and collaborated with search engine providers to remove the exposed data. CEO Sam Altman acknowledged the incident, emphasizing concerns regarding AI's handling of sensitive user information often shared in confidence.

Key employees

Board of directors of the OpenAI nonprofit

  • Bret Taylor (chair), former chairman of Twitter’s board and co-CEO of Salesforce
  • Sam Altman
  • Adam D'Angelo, co-founder and CEO of Quora
  • Sue Desmond-Hellmann, previous CEO of the Bill & Melinda Gates Foundation
  • Nicole Seligman, attorney and former Executive Vice President at Sony Corporation
  • Paul Nakasone, served as Director of the National Security Agency from 2018 to 2024
  • Zico Kolter, expert in computer science
  • Adebayo Ogunlesi, managing partner at Global Infrastructure Partners

Key individual investors

  • Reid Hoffman, co-founder of LinkedIn
  • Peter Thiel, co-founder of PayPal
  • Jessica Livingston, founding partner at Y Combinator
  • Elon Musk, co-founder

Changes in personnel

In 2018, Elon Musk stepped down from OpenAI’s Board of Directors, citing a potential conflict of interest due to Tesla’s AI work on autonomous vehicles, where he serves as CEO. His financial backing for OpenAI was reported to be under $45 million.

On March 3, 2023, Reid Hoffman exited the board to prevent conflicts related to his AI ventures through Greylock Partners and his startup Inflection AI, while maintaining his position on Microsoft’s board, a significant investor in OpenAI.

May 2024 saw Chief Scientist Ilya Sutskever resigning, with Jakub Pachocki appointed as his replacement. Co-leader Jan Leike also left amid concerns regarding safety and trust. Subsequent to these shifts, OpenAI established collaborations with Reddit, News Corp, Axios, and Vox Media. Shortly thereafter, Paul Nakasone joined the board.

In August 2024, co-founder John Schulman departed to join Anthropic, and president Greg Brockman took an extended leave until November. The following month, CTO Mira Murati also announced her departure.

By November 2025, Lawrence Summers resigned from the board amid controversy surrounding leaked emails.

Corporate governance and legal matters

In May 2023, Sam Altman, Greg Brockman, and Ilya Sutskever presented governance proposals addressing superintelligence, anticipating its arrival within a decade that could unlock significant prosperity but also existential threats. They recommended creating an international regulatory body similar to the IAEA to supervise AI systems surpassing a defined capability level while advocating lighter oversight for less advanced AI. Additionally, they emphasized the need for expanded technical safety research alongside coordinated governmental initiatives integrating existing projects.

In July 2023, the FTC issued a civil investigative demand to OpenAI investigating whether its data security and privacy protocols during ChatGPT development violated Section 5 of the Federal Trade Commission Act of 1914, potentially causing consumer harm including reputational damage. Although such inquiries are usually confidential, the FTC document was leaked. The FTC requested comprehensive information about OpenAI’s technologies, privacy protections, and efforts to prevent false or defamatory chatbot responses. Concerns were also raised about reciprocal expenditures between Microsoft and OpenAI involving credits for Microsoft Azure and shared engineering resources, highlighting possible public harm and antitrust implications.

In September 2024, OpenAI’s global affairs lead voiced endorsement for the UK's "smart" approach to AI regulation during testimony before a House of Lords committee.

By February 2025, CEO Sam Altman expressed OpenAI’s willingness to engage with China despite U.S. regulatory constraints. This stance followed market disruption caused by Chinese AI company DeepSeek, which released open models such as DeepSeek V3 and R1. In response, OpenAI intensified its security measures to combat industrial espionage amid accusations that DeepSeek appropriated its distillation methods.

As of March 2025, the United States had enacted 781 state-level AI laws. OpenAI advocated for federal preemption to harmonize AI regulation nationwide, opposing California’s AI legislation as overstepping federal jurisdiction. Conversely, organizations like Public Citizen cautioned that federal preemption might hinder innovation, citing OpenAI’s own expansion under existing state regulations.

Non-disparagement agreements

Prior to May 2025, OpenAI mandated that departing staff sign perpetual non-disparagement agreements which prohibited them from criticizing the company and required acknowledgment of the agreement itself. Researcher Daniel Kokotajlo, a former employee, publicly revealed that he relinquished his vested equity in order to leave without consenting to these terms. Sam Altman stated he was unaware of any clause that would cancel vested equity and asserted that OpenAI never applied such provisions to revoke equity. Nevertheless, leaked emails and documents appear to contradict these claims. On May 23, 2024, OpenAI circulated an internal memo officially releasing ex-employees from these non-disparagement obligations.

Copyright

Starting in July 2023, OpenAI became the target of several copyright infringement lawsuits initiated by writers including Sarah Silverman, Matthew Butterick, Paul Tremblay, and Mona Awad. By September 2023, a group of 17 leading authors such as George R. R. Martin, John Grisham, Jodi Picoult, and Jonathan Franzen joined the Authors Guild in filing a class action suit accusing OpenAI of unauthorized use of their copyrighted material. Later that year, The New York Times also pursued legal action against OpenAI on similar grounds. In May 2024, OpenAI disclosed it had deleted its Books1 and Books2 datasets—collections utilized during GPT-3’s training and believed by the Authors Guild to include over 100,000 copyrighted books.

In 2021, OpenAI created Whisper, a speech-to-text system employed to transcribe more than one million hours of YouTube content for GPT-4’s training purposes. This project raised internal concerns about potential breaches of YouTube’s terms of service; nevertheless, it advanced with significant oversight from OpenAI president Greg Brockman. The resulting dataset was instrumental in developing GPT-4.

Early in 2024, digital news organizations such as The Intercept, Raw Story, and Alternate Media Inc. initiated lawsuits against OpenAI concerning copyright violations linked to AI training datasets. This represented a novel legal strategy by digital-only publishers. On April 30, 2024, eight newspapers—including The Mercury News, Chicago Tribune, and New York Daily News—filed suit against both OpenAI and Microsoft for unauthorized reproduction of their copyrighted articles.

A lawsuit filed in June 2023 accused OpenAI of scraping approximately 300 billion words from the internet without permission or registering as a data broker. The plaintiffs alleged that OpenAI and Microsoft unlawfully gathered personal data worldwide to train AI models.

On May 22, 2024, OpenAI announced a content licensing agreement with News Corp to integrate news from outlets such as The Wall Street Journal, New York Post, The Times, and The Sunday Times into its AI platform. Meanwhile, other media companies like The New York Times continued pursuing litigation over copyright infringement related to AI training data usage. In November 2024, prominent Canadian news organizations including Toronto Star and CBC filed lawsuits against OpenAI for unauthorized use of their published articles.

In October 2024, former OpenAI engineer Suchir Balaji publicly accused the company of infringing copyrights during the development of commercial large language models (LLMs). Identified as a potential critical witness in ongoing legal proceedings, Balaji died by suicide in November 2024. His passing triggered conspiracy theories and prompted calls for investigation from California Congressman Ro Khanna and others.

GDPR compliance

In April 2023, the European Data Protection Board (EDPB) of the EU established a dedicated task force on ChatGPT to coordinate regulatory enforcement efforts following actions taken by Italy’s data protection authority against OpenAI concerning ChatGPT.

By April 2024, NOYB submitted a formal complaint to Austria’s Datenschutzbehörde accusing OpenAI of breaching the European General Data Protection Regulation. The grievance highlighted an instance where ChatGPT generated an incorrect birthdate for a living individual without providing that person access to or the ability to amend their personal information. Furthermore, OpenAI indicated it was unable to reveal recipients of ChatGPT-generated content or the underlying data sources.

Military and warfare

OpenAI encountered backlash after removing its prohibition on employing ChatGPT for "military and warfare" applications. Prior to January 10, 2024, its usage guidelines explicitly banned "weapons development" alongside "military and warfare" activities. The revised policy now broadly prohibits using the platform to inflict harm on oneself or others or to create or deploy weapons. Additionally, OpenAI is reportedly partnering with the United States military on cybersecurity initiatives.

Wrongful-death lawsuits over ChatGPT safety (2025)

In August 2025, the parents of a 16-year-old who died by suicide initiated a wrongful death lawsuit against OpenAI and CEO Sam Altman. They contended that extended interactions with ChatGPT regarding mental health and self-harm contributed to their child’s death and claimed that protections for minors were insufficient. OpenAI offered condolences and announced strengthened safety measures, including enhanced crisis response protocols and parental controls. This case represented an unprecedented wrongful death claim targeting the company’s chatbot technology. The lawsuit was filed in a California state court located in San Francisco.

In November 2025, the Social Media Victims Law Center along with the Tech Justice Law Project launched seven lawsuits against OpenAI, four of which alleged wrongful death. These legal actions involved individuals who died by suicide following prolonged use of ChatGPT: Zane Shamblin (23, Texas), Amaurie Lacey (17, Georgia), Joshua Enneking (26, Florida), and Joe Ceccanti (48, Oregon). The complaints assert that emotional manipulation and inadequate safety protocols played roles in these fatalities.

Retrieved from "https://www.alfawiki.org/index.php?title=OpenAI&oldid=68457"