OpenAI may leave California if regulators block its for-profit restructuring

OpenAI is warning that it might leave California entirely as legal fights and political backlash threaten to kill its restructuring plans.

Executives are weighing a possible exit after realizing the state’s attorney general could block the company’s move to become a for-profit operation.

The company’s entire future now hinges on getting regulatory approval, or risking losing nearly $19 billion in investor cash.

According to reporting from The Wall Street Journal, OpenAI is being targeted by some of California’s biggest nonprofit coalitions, labor unions, and philanthropic groups. These organizations want the attorney general to investigate whether the company’s restructuring would break state charity laws.

Attorneys General threaten legal action if restructure proceeds

California and Delaware’s attorneys general are both investigating the proposal. They’ve got legal power to sue or demand changes if they believe OpenAI is violating nonprofit law. The office of California’s attorney general is already warning that OpenAI’s current plan may go against its original mission.

The state sent a letter raising concerns, especially in light of multiple suicides reported by people who interacted with ChatGPT over extended periods.

In the letter, regulators wrote, “The recent deaths are unacceptable. They have rightly shaken the American public’s confidence in OpenAI and this industry.” They told OpenAI that safety must come first.

They also made it clear the company’s nonprofit status requires transparency and a public-first approach to AI deployment. Executives inside OpenAI didn’t expect this kind of pushback when they announced the restructure late last year.

The intensity of the legal pressure, especially from California, has turned into a real threat. Sam Altman, who co-founded OpenAI in 2015, built the company into an $86 billion AI powerhouse, but now he’s at risk of watching it fall apart over legal technicalities.

OpenAI is still run as a for-profit unit under a nonprofit parent, and that structure doesn’t sit well with big investors. They’re not getting traditional shares and want the change fast.

Those investors have already promised billions, but only if the new company can legally issue stock. Without the restructure, OpenAI could lose the cash it needs to stay in the AI arms race.

Projects like building custom chips, setting up new data centers, and fighting off poaching from rivals are all on the line. There’s also pressure to reduce huge yearly losses and wrestle back more control from Microsoft, which has a massive stake in the business.

OpenAI hires political allies and concedes to critics

OpenAI has been scrambling for political support. The company hired several advisers tied to California Governor Gavin Newsom, including former Senator Laphonza Butler.

They’ve spent the summer hosting closed-door meetings with advocacy groups across the state and promised to pour $50 million into nonprofits and communities.

In May, the company made a key concession. Instead of separating the nonprofit and for-profit sides, executives agreed the nonprofit would stay in charge of the new company. Internally, that was seen as a blow for Sam and his investors, but it was the only way to calm some of the opposition.

Despite the changes, the heat hasn’t gone down. Meta has asked California officials to stop the restructuring. Elon Musk offered to take control of OpenAI’s assets earlier this year but got rejected. 

He’s now suing the company through his rival AI startup xAI, saying the restructure breaks the original nonprofit agreement. The trial is scheduled for next year, and OpenAI says the lawsuit is “baseless.”

In April, a group of more than 60 nonprofits led by the San Francisco Foundation demanded an investigation into whether OpenAI has violated its federal tax-exempt status.

In a letter, they warned that without enforcement, more tech startups might abuse nonprofit structures to benefit private investors. “Other startups, looking at OpenAI as a model, are likely to consider whether to take similar advantage,” they wrote.

Meanwhile, OpenAI is trying to fix public concerns. Bret Taylor, chairman of the board, said they’re working on adding parental controls to ChatGPT and tackling “sycophancy,” a flaw where the AI agrees too easily with users. “We are fully committed to addressing the Attorneys General’s concerns,” Bret said.

Want your project in front of crypto’s top minds? Feature it in our next industry report, where data meets impact.