OpenAI officially announced a major adjustment late at night! Letter from all Ultraman employees: Give up profit-making transformation and open source powerful models
Updated on:June-29th-2025
Recommendation
OpenAI subverts tradition and transforms its for-profit business into a public interest company, committed to making AI benefit all mankind.
Core content:
1. OpenAI transforms into a "public interest company" to pursue a balance between public interest and shareholder interests
2. Non-profit organizations continue to control OpenAI to protect its mission and original intention
3. Sam Altman emphasizes OpenAI's unique mission orientation to ensure that AGI benefits all mankind
Yang Fangxian
Founder of 53AI/Most Valuable Expert of Tencent Cloud (TVP)
Ten years ago, OpenAI was born on a kitchen table. At that time, the founders were not thinking about how to make profits, but how to make AI serve humanity.Ten years later, this technology giant, which is valued at hundreds of billions and owns ChatGPT, is still asking the original question: To whom should AGI belong and for whom should it be used?Early this morning, OpenAI’s board of directors and founder Sam Altman gave an institutional answer in an open letter: its for-profit business will be transformed into a “public benefit company” (PBC). The mission will remain unchanged and the non-profit organization will continue to control it, but the governance framework will be clearer.Similar to the structure currently adopted by companies such as Anthropic and xAI, the public benefit company is a special form of company that adds the pursuit and responsibility of public interest to the traditional company's pursuit of maximizing shareholder interests.If AGI is the new engine of an industrial revolution, OpenAI would obviously prefer to hold the steering wheel first rather than set off immediately.OpenAI also summarized the four key points of this organizational structure adjustment on the X platform:OpenAI will continue to be controlled by its existing nonprofit organizationOur existing for-profit entity will be transformed into a Public Benefit Corporation (PBC)The nonprofit will take control of PBC and become its major shareholderThe nonprofit and PBC will continue to have the same missionHowever, after the "palace fight" in 2023 and Ilya's subsequent resignation, OpenAI's original intention is still widely questioned by the outside world. Especially in the face of the collapse of Altman's personality, netizens are no longer easily convinced.There was no shortage of pointed questions in the comments section: "Do nonprofit organizations really have more than 51% of the control?"Letter from the Board of Directors to EmployeesThe OpenAI Board of Directors has developed an updated plan to drive the evolution of OpenAI’s organizational structure.OpenAI was originally founded as a nonprofit and is still supervised and controlled by that nonprofit today. It will continue to be supervised and controlled by that nonprofit going forward.Our for-profit limited liability company (LLC), which has been affiliated with the nonprofit since 2019, will now transform into a public benefit corporation (PBC)—a mission-driven corporate structure that takes into account both shareholder interests and mission.The nonprofit will take control of PBC and become one of its major shareholders, giving it greater resources to support multiple public interests.Our mission remains the same and the PBC will uphold the same mission.After listening to civic leaders and engaging in constructive dialogue with the Delaware and California Attorneys General’s Offices, we have decided that OpenAI will continue to be controlled by a nonprofit organization. We are grateful for the support of both Attorneys General’s Offices and look forward to continuing these important conversations to ensure that OpenAI can effectively advance its mission of making AGI a reality for all of humanity.Below is a letter Sam wrote to employees and stakeholders explaining why we are so excited about this new direction.Sam's letter to employeesOpenAI is not an ordinary company, and it never will be.Our mission is to ensure that artificial general intelligence (AGI) benefits all of humanity.When we started OpenAI, we had no idea how we were going to accomplish this mission. We were just sitting around our kitchen table thinking about what research to do.We hadn’t thought about products or business models at the time. We couldn’t imagine the direct benefits of AI, such as providing medical advice, improving learning and productivity, and we didn’t anticipate that training models and serving users would require hundreds of billions of dollars in computing power.We really don’t know how AGI will be built or how it will be used. Many people imagine AGI to be like an “oracle” that can tell scientists and the president what to do, and while this may be extremely dangerous, perhaps a few people can be trusted to control it.In the early days of OpenAI, many believed that AI should only be in the hands of a few trusted people.But now we see a new path—AGI can become the most powerful tool in human history, directly empowering every single person.If we can achieve this, we believe people will create amazing things for each other, driving progress in society and quality of life.Of course, it won’t be used only for good purposes, but we believe that human kindness will far outweigh the potential negative impacts.We are firmly on the path of "democratizing AI". We want to put powerful tools in the hands of everyone. We are surprised and delighted by the results created by users with our tools, and we also see their enthusiasm for using it.We want to open source some very powerful models. We want to give users a high degree of freedom to use these tools within broad boundaries - even if their values don't completely align with ours - and let users decide how ChatGPT behaves.We believe this is the best path forward - that AGI should be a means for all of humanity to help each other.We understand that not everyone agrees with this view.We want to build a brain that belongs to all of humanity and make it easy for everyone to use it to do what they want (as long as it doesn't infringe on the freedom of others).People are using ChatGPT to improve their productivity as scientists, programmers, etc.People are also using it to solve major medical challenges and learn more.People are also using it to get advice on how to deal with complex situations.We are proud to offer a service that truly helps countless people and is a direct reflection of our mission.But people want to use more; and our current supply of AI is far behind demand, forcing us to throttle systems and slow them down.As the system's capabilities continue to increase, users will want to do more and more exciting things with it.When we founded this research lab nearly a decade ago, we had no idea the world would be where it is today, but we couldn’t be more excited to see where it is now.It was time to change our organizational structure. We wanted to achieve three things:We want to operate in a way that gets us the resources we need to make our services broadly accessible to all of humanity. This will require hundreds of billions of dollars today, and potentially trillions of dollars in the future. We believe this is the best way to achieve our mission and the path for people to create tremendous value for each other with these new tools.We want our nonprofit to be the largest, most effective nonprofit in history, focused on using AI to achieve the most leveraged positive outcomes.We want to deliver beneficial AGI. This includes leading the way in safety and alignment; we are proud of the safety record of our current released systems, as well as our alignment research, "red team testing" process, and transparent mechanisms like model specifications. As AI progresses at an accelerated pace, our commitment to safety is even stronger. We want democratized AI to win.After listening to civic leaders and discussing the issues with the California and Delaware Attorneys General’s Offices, we have decided that the nonprofit organization will continue to own the company and we look forward to working with them, Microsoft, and the newly appointed nonprofit board members to further advance this plan.OpenAI was originally founded as a nonprofit, and it remains a nonprofit organization that oversees and controls the for-profit portion, and it will continue to be structured that way. That will not change.The for-profit LLC under the nonprofit organization will be converted into a public benefit corporation (PBC), with the mission remaining the same.Many other AGI labs (such as Anthropic and X.ai), as well as mission-driven companies like Patagonia, have adopted the PBC structure, and we think it will work for us as well.We will abandon the current complex "profit cap structure" - it made sense in the world of one dominant AGI project, but no longer makes sense in today's world of multiple excellent AGI companies. We will adopt a conventional equity structure, with everyone holding shares. This is not a sale, but a move to a simpler structure.The nonprofit will continue to control PBC and become a significant shareholder in PBC, with a shareholding ratio supported by an independent financial advisor. This will provide the nonprofit with more resources to support AI in bringing benefits to different communities and continue to advance our mission.As PBC grows, the nonprofit's resources increase, allowing it to do more work.We look forward to receiving upcoming recommendations from the Nonprofit Commission on how to ensure that AI truly benefits everyone, not just a few.Their recommendations will focus on how nonprofit efforts can advance a more democratic AI future and have tangible impact in areas such as health, education, public services, and scientific discovery.We believe this structure will allow us to continue to move quickly and safely to bring powerful AI to everyone.Building AGI is one brick we’re laying for human progress, and we can’t wait to see what you’ll add next.