OpenAI wants to break away from its nonprofit roots and transform into a for-profit company

OpenAI’s history as a nonprofit research institute that also sells commercial products like ChatGPT may be coming to an end as the San Francisco company looks to more fully transform itself into a for-profit company accountable to its shareholders.

The artificial intelligence company’s board is considering a decision that would convert it to a nonprofit corporation, according to a source familiar with the discussions who was not authorized to speak publicly.

While OpenAI already has a for-profit division where most of its employees work, it is controlled by a nonprofit board whose mission is to help humanity. That would change if the company transforms the core of its structure into a non-profit corporation, a type of corporate entity that is intended to both help society and generate profits.

The board has not yet made a final decision and the timing of the postponement has not yet been determined, the source said.

OpenAI CEO Sam Altman acknowledged in public statements Thursday that the company was considering a restructuring, but said the departures of key executives the day before were unrelated.

Speaking at a technology conference in Italy, Altman mentioned that OpenAI had been considering an overhaul to move to the “next level.” But he said it had nothing to do with the resignations of technology chief Mira Murati and two other top executives on Wednesday.

“OpenAI will be stronger in this, as we are in all our transitions,” Altman said at the Italian Tech Week event in Turin. “I’ve seen a few things related to restructuring. That is absolutely not true. Most of the things I saw were also completely wrong,” he said without elaborating.

“But we have thought about restructuring,” he added.

OpenAI said Thursday that it will continue to maintain a nonprofit arm.

“We remain focused on developing AI that benefits everyone, and as we have previously shared, we are working with our board to ensure we are best positioned to succeed in our mission,” one said written declaration. “The nonprofit is core to our mission and will continue to exist.”

The resignations of Murati, Chief Research Officer Bob McGrew and another research leader, Barret Zoph, were “all about people being ready for new chapters of their lives and a new generation of leaders,” Altman said.

The departures were the latest in a string of recent high-profile departures, which also includes the resignations of OpenAI co-founder Ilya Sutskever and head of security Jan Leike in May. In a statement, Leike criticized OpenAI because security had “moved into the background of shiny products.”

Much of the conflict at OpenAI stems from its unusual governance structure. The company was founded in 2015 as a non-profit organization with the goal of safely developing futuristic AI to help humanity. Today it is a rapidly growing large corporation still controlled by a nonprofit board committed to its original mission.

This unique structure allowed four OpenAI board members – Sutskever, two outside tech entrepreneurs and an academic – to briefly oust Altman last November in what was later described as a dispute over a “significant loss of trust” between the board and top management. Management. But with the help of a powerful backer, Microsoft, Altman was brought back to the CEO position days later and a new board replaced the old one. OpenAI also reappointed Altman to the board in March.

It may not be easy to change OpenAI’s corporate structure, even if it is designed to keep investors and employees happy.

Tax experts said OpenAI’s corporate structure appeared to be designed to give the tax-exempt nonprofit organization full control over the for-profit entities the organization created as it grew.

In 2016, the goal of the founders of OpenAI – a group that included Altman and Tesla CEO Elon Musk – was to “advance digital intelligence in a way that is most likely to benefit humanity as a whole, without necessity to generate financial returns, to be restricted.”

A few years later, the organization realized that it needed more computing power to continue developing AI technologies and that it would need to raise billions of dollars to do so. “We want to increase our ability to raise capital while serving our mission, and no pre-existing legal structure that we are aware of provides the right balance,” co-founders Sutskever and Greg Brockman wrote in a 2019 blog post.

So they created a new for-profit company with a “cap” on the amount of profits investors or employees could make, and put the nonprofit and its board in charge of running the new company.

Any “excess” profits would go back to the nonprofit, Brockman and Sutskever explained, although in practice little money has gone back to the nonprofit in recent years. Brockman has been on leave since August, leaving Altman one of the few early leaders still at the helm.

In a study published in February, Ellen P. Aprill, professor emeritus of tax law at LMU’s Loyola Law School, examined OpenAI’s corporate structure and found that it appeared to be “carefully” designed to protect its nonprofit status.

All of its subsidiaries are directed or managed by the nonprofit and its board, and OpenAI says it warns investors that they may never receive a return.

But April and her colleagues pointed out that Altman’s firing and reinstatement was evidence that the nonprofit’s board may not have significant responsibility. “Unless board members fulfill their fiduciary duties … even the most careful structures are for naught,” April and her co-authors wrote.

——

The Associated Press and OpenAI have a license and technology agreement that allows OpenAI to access a portion of AP’s text archives.