OpenAI

OpenAI is an American artificial-intelligence research and deployment organization headquartered in San Francisco. It was established in December 2015 as a nonprofit research institute with a mission to ensure that artificial general intelligence benefits all of humanity. The founding group included technology entrepreneur Elon Musk, investor and entrepreneur Sam Altman, researcher Ilya Sutskever, and several prominent machine-learning engineers and scientists. Its original public identity emphasized open research, broad cooperation, and concern about the potential benefits and risks of highly capable AI. The organization initially concentrated on foundational research rather than consumer software. Early projects included OpenAI Gym, a platform for reinforcement-learning research, and Universe, an environment intended to test AI systems across games, websites, and other applications. The company also developed systems capable of playing complex games, including Dota 2, as demonstrations of reinforcement learning and increasingly general-purpose machine intelligence. OpenAI changed its structure in 2019. It created a capped-profit subsidiary to attract investment, issue equity to employees, and fund the substantial computing costs required to train large models. Microsoft became its principal strategic technology partner, providing capital and Azure computing infrastructure. The partnership helped OpenAI scale the GPT family and commercialize access through an application programming interface and products built on top of its models. The resulting structure retained the nonprofit entity as the controlling organization while allowing the operating business to raise private capital. The GPT series became OpenAI’s central technology platform. GPT-3 demonstrated the commercial potential of large language models, while GPT-4 expanded capabilities in reasoning, writing, coding, image understanding, and other tasks. OpenAI also developed DALL-E, a family of text-to-image systems, and related multimodal technologies. The launch of ChatGPT in November 2022 transformed the company from a specialist research organization into a globally recognized consumer and enterprise technology brand. ChatGPT popularized conversational generative AI and contributed to a broader surge of investment, competition, and public debate around the technology. OpenAI’s growth has been accompanied by governance, legal, safety, and commercial controversies. In November 2023, its board briefly removed Sam Altman as chief executive, citing a loss of confidence, before he returned several days later after negotiations involving employees, investors, and Microsoft. The episode led to changes in the board and intensified debate about the relationship between the nonprofit mission and the commercial business. The company has also faced copyright litigation from authors, newspapers, and other rights holders concerning the use of published material in training datasets. Several safety researchers departed during 2024 amid concerns that commercial priorities were reducing the emphasis placed on long-term AI safety. OpenAI continues to operate internationally through consumer subscriptions, enterprise offerings, developer APIs, research programs, and partnerships. Its business spans large language models, image generation, speech and multimodal systems, developer tools, productivity applications, and infrastructure initiatives. The organization remains privately held and is controlled through a structure involving the nonprofit OpenAI Foundation and the for-profit OpenAI Group public benefit corporation.

OpenAI was founded in Delaware in December 2015 as a nonprofit artificial-intelligence research organization. Its founding group included Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman, John Schulman, Wojciech Zaremba, and other researchers and technology figures. The founders framed the project as a long-term effort to develop advanced AI in a way that would benefit humanity. They were particularly concerned that systems approaching or exceeding human-level intelligence could create both extraordinary social value and severe risks if developed or deployed irresponsibly. The organization began with an unusually open research orientation. It sought to collaborate with universities and other laboratories and initially stated that research and patents would be shared when appropriate. Its early work focused on reinforcement learning and general-purpose AI research. OpenAI Gym, released publicly in 2016, gave researchers a common environment for evaluating reinforcement-learning algorithms. Later that year, Universe extended the idea by providing a framework for testing AI agents across games, websites, and software applications. OpenAI also developed increasingly sophisticated game-playing systems, culminating in widely publicized Dota 2 demonstrations. As model training became more computationally demanding, OpenAI’s operating model changed. In 2019, it created a capped-profit structure and a commercial subsidiary. This allowed the organization to obtain large-scale private investment, offer equity to employees, and purchase the computing resources needed for frontier-model development. Microsoft invested substantially and became OpenAI’s primary cloud partner through Azure. The arrangement connected OpenAI’s research to a global commercial distribution network and helped turn its models into APIs and products for developers and businesses. The GPT family became the organization’s principal research and product platform. GPT-3 showed that very large language models could perform a wide range of writing, question-answering, summarization, and programming tasks from natural-language instructions. OpenAI subsequently developed GPT-4 and other multimodal systems, while DALL-E extended the company’s work into image generation. The November 2022 release of ChatGPT made these capabilities accessible through a simple conversational interface. Its rapid adoption catalyzed widespread interest in generative AI, prompted major investments across the technology industry, and made OpenAI one of the most recognizable AI brands worldwide. OpenAI’s expansion generated significant institutional tension. The November 2023 removal of Sam Altman by the board exposed disagreements about governance, communication, and the balance between rapid commercialization and the organization’s founding mission. Altman returned after several days, and the board was restructured. During 2024, departures by senior leaders and safety researchers produced further criticism that the company’s safety function was being weakened or subordinated to product and revenue goals. OpenAI also faced copyright lawsuits from authors and media organizations concerning material allegedly used to train its models. The organization continued to broaden its business through ChatGPT subscriptions, enterprise services, developer APIs, image and audio products, and partnerships. Its capital requirements grew as model training and inference demanded increasingly large data-center and cloud investments. In 2025, OpenAI adopted a public benefit corporation structure for its commercial arm, with the nonprofit entity renamed the OpenAI Foundation. The resulting arrangement was intended to combine access to conventional investment with formal mission-oriented governance. OpenAI remains privately held and continues to present itself as both a frontier research institution and a large-scale AI product company.

1. 2025Commercial operation becomes a public benefit corporation

   OpenAI reorganizes its commercial business as OpenAI Group PBC and establishes the OpenAI Foundation as the nonprofit entity.

   Source
2. 2023GPT-4 is announced

   OpenAI announces GPT-4, a more capable model supporting text and image inputs in selected contexts.

   Source
3. 2023Leadership crisis and board reconstruction

   Sam Altman is removed and then reinstated as CEO, followed by a reconstruction of OpenAI’s board.

   Source
4. 2022ChatGPT launches

   OpenAI releases ChatGPT, making conversational generative AI accessible to a broad public audience.

   Source
5. 2020GPT-3 API is introduced

   OpenAI begins commercializing GPT-3 capabilities through an API for developers and businesses.

   Source
6. 2019Capped-profit structure is created

   OpenAI forms a capped-profit subsidiary to attract investment and finance large-scale model training while retaining nonprofit control.

   Source
7. 2016OpenAI Gym is released

   The organization releases OpenAI Gym, a public platform for developing and comparing reinforcement-learning systems.

   Source
8. 2016Universe research platform launches

   OpenAI introduces Universe, a framework for testing AI agents across games, websites, and other software environments.

   Source
9. 2015OpenAI is founded as a nonprofit

   OpenAI is established in Delaware as a nonprofit AI research organization focused on ensuring that artificial general intelligence benefits humanity.

   Source

A frontier AI research and deployment organization combining advanced model development, consumer applications, enterprise software, and developer infrastructure while maintaining a stated mission centered on broadly beneficial artificial general intelligence.

• ChatGPTConversational AI assistant2022

  ChatGPT is OpenAI’s conversational assistant and best-known consumer product. It uses generative language models to answer questions, draft and transform text, assist with coding, summarize information, and support multimodal interactions. The service is offered through consumer and business plans and serves as an interface for deploying successive OpenAI model generations.

• GPTLarge language and multimodal models2018

  The GPT family comprises OpenAI’s general-purpose generative models. They are used for language generation, reasoning, coding, analysis, and, in later generations, combinations of text, image, audio, and other modalities. GPT models power ChatGPT, the OpenAI API, and integrations developed by external organizations.

• DALL-EText-to-image generation2021

  DALL-E is OpenAI’s image-generation product family. It creates and edits images from natural-language prompts and has been integrated with ChatGPT and other OpenAI interfaces. The product expanded OpenAI’s offering beyond language models into visual content creation.

• OpenAI APIDeveloper platform2020

  The OpenAI API gives developers programmatic access to selected OpenAI models and capabilities. Customers use it to build software for text generation, embeddings, image creation, speech, classification, automation, and other AI-enabled workflows.

• OpenAI GymAI research platform2016

  OpenAI Gym is a toolkit and collection of environments for reinforcement-learning research. It helped standardize experimentation and benchmarking for researchers developing agents that learn through interaction.

• UniverseAI research platform2016

  Universe was a research platform designed to evaluate and train AI agents across a broad range of software environments, including games and websites. It represented OpenAI’s early interest in measuring general-purpose competence.

• 2023ChatGPT Plus and enterprise expansion

  Global

  OpenAI expanded ChatGPT from a primarily public research preview into a tiered commercial service, adding paid access and enterprise-oriented offerings with stronger administrative and privacy features.

  Outcome. The strategy created recurring subscription and enterprise channels alongside API revenue.

  Source
• 2022ChatGPT public launch

  Global

  OpenAI introduced ChatGPT as a broadly accessible conversational AI service. The low-friction interface encouraged experimentation by consumers, educators, developers, and businesses and helped establish generative AI as a mainstream technology category.

  Outcome. The launch generated exceptional public adoption and made OpenAI a leading global AI brand.

  Source

• 2025Convert the commercial arm into a public benefit corporationStrategy

  OpenAI sought a structure capable of attracting conventional investment while preserving a formal connection to its public-benefit mission.

  What changed. The commercial operation was reorganized as OpenAI Group PBC, while the nonprofit entity became the OpenAI Foundation.

  Aftermath. The reorganization preserved nonprofit participation in governance but also generated legal and public criticism concerning mission protection and control.

  Source
• 2023Launch GPT-4Product launch

  OpenAI sought to improve the reliability, reasoning, coding, and multimodal capabilities of its large language models.

  What changed. OpenAI released GPT-4 and made access available through selected products and its developer platform.

  Aftermath. GPT-4 became a major foundation for ChatGPT, enterprise deployments, and third-party applications.

  Source
• 2022Release ChatGPTProduct launch

  OpenAI had developed increasingly capable language models but had not yet offered them through a mass-market conversational interface.

  What changed. The company launched ChatGPT as a public conversational service based on its generative language-model technology.

  Aftermath. ChatGPT rapidly expanded public awareness of generative AI and accelerated competition across the technology industry.

  • Microsoft — Microsoft expanded the integration of OpenAI models into Bing, Windows, and Copilot products.
  • Google — Google accelerated development and deployment of its own conversational generative-AI products.
  Source
• 2019Create a capped-profit operating structureStrategy

  Frontier-model research required substantially more capital and computing capacity than the original nonprofit structure could easily provide.

  What changed. OpenAI created a capped-profit subsidiary, issued an investment structure, and entered a major partnership with Microsoft while retaining nonprofit control.

  Aftermath. The structure enabled large-scale investment and commercialization but created continuing debate about mission alignment and governance.

  Source

• 2024Debate over AI safety prioritiesControversy

  The departure of safety researchers and leaders prompted former employees and outside critics to argue that OpenAI was placing product development and commercial growth ahead of long-term safety work.

  Source
• 2023Sam Altman leadership crisisControversy

  OpenAI’s board removed Sam Altman as CEO, citing a loss of confidence in his communications. Employees and investors challenged the decision, and Altman returned after several days following negotiations and a board reconfiguration. The episode exposed tensions over governance and the company’s nonprofit mission.

  Source
• 2023Copyright litigation over AI trainingControversy

  Authors, publishers, and media companies filed lawsuits alleging that copyrighted works were used to train OpenAI systems without permission. The disputes concern fair use, licensing, attribution, and the legal status of model training and generated outputs.

  Source

• 2025OpenAI introduces its public benefit corporation structure

  OpenAI reorganized its commercial operation as a public benefit corporation while retaining a nonprofit foundation intended to preserve mission-related oversight.

  OtherLeadership change
  Source
• 2024OpenAI completes major private financing

  OpenAI completed a private financing round that supported continued investment in computing capacity, model research, and commercial expansion.

  Other
  Source
• 2023Microsoft expands its strategic investment in OpenAI

  Microsoft announced a multiyear investment reported at approximately $10 billion and continued to provide Azure computing resources and model distribution through products such as Copilot.

  M&AOther
  Source
• 2023OpenAI announces GPT-4

  OpenAI introduced GPT-4 as a more capable multimodal model with improved performance across language, reasoning, coding, and image-related tasks.

  Product launchProduct generation
  Source
• 2023OpenAI launches DALL-E 3

  DALL-E 3 improved text-to-image generation and was integrated with ChatGPT, allowing users to describe and refine image concepts through natural-language interaction.

  Product launchProduct generation
  Source
• 2022OpenAI launches ChatGPT

  OpenAI released ChatGPT, a conversational interface built on its generative language models. The product rapidly broadened public awareness of generative AI and became the company’s most prominent consumer offering.

  Product launchCampaign
  Source

1. OpenAI
2. OpenAI official website
3. OpenAI: ChatGPT launch announcement
4. OpenAI: GPT-4 announcement
5. Microsoft and OpenAI extend partnership
6. OpenAI copyright lawsuits
7. 2023 OpenAI leadership crisis
8. OpenAI Gym is released
9. Universe research platform launches
10. GPT-3 API is introduced
11. ChatGPT Plus and enterprise expansion
12. OpenAI完成重大私人融资
13. OpenAI推出DALL-E 3