OpenAI Backtracks on Plans
OpenAI said on Monday that it was restructuring as a public benefit corporation, allowing the nonprofit that controls OpenAI Backtracks
to retain its grip on the company.
The decision is a victory for OpenAI’s critics, including one of its founders, Elon Musk, who complained that the company was too focused on profits and had abandoned its early plan to build artificial intelligence systems with safety foremost in mind.
The changes announced on Monday are the latest in years of corporate drama for what many consider to be the most influential A.I. company in the world. OpenAI’s ChatGPT, released in late 2022, was an overnight success that sent the rest of the tech industry scrambling. In just a few years, tech’s biggest companies have spent billions on their own A.I. projects, with hundreds of billions more planned for this decade.
Mr. Musk, who is now running his own A.I. company, sued OpenAI over plans it was putting into place to change its corporate structure from an unorthodox system that gave a nonprofit oversight of a for-profit company. But he was not the only critic of OpenAI’s planned changes. The attorneys general in California, where OpenAI is headquartered, and in Delaware, where it was legally created, also said that they were monitoring its restructuring. The office of the California attorney general, Rob Bonta, said in a statement that it was reviewing OpenAI’s new plan.
And in recent weeks, a number of academics from the legal community and experts such as Geoffrey Hinton, who won a Nobel Prize last year for his pioneering A.I. research, also publicly expressed concern about OpenAI’s direction.
The argument over how OpenAI should be structured and what its priorities should be homed in on a fundamental question about artificial intelligence: Should researchers rush headlong to develop new and more powerful A.I. systems? Or should the theoretical risk that A.I. presents to humanity inform everything those researchers create?
OpenAI was started in 2015 with that tension in mind.
Sam Altman, OpenAI chief executive, created the artificial intelligence organization with several other Silicon Valley figures as a nonprofit in late 2015. In 2018, after Mr. Musk left in a power struggle, Mr. Altman attached OpenAI to a for-profit company so he could raise the billions of dollars needed to build A.I. technologies.
But the nonprofit retained its grip in a structure that some saw as an albatross to the company’s growth. Last year, Mr. Altman and his company began working on a plan to shift control from the nonprofit to OpenAI’s investors, so that it would be more attractive to them.
Soon after, Mr. Musk sued OpenAI, Mr. Altman and another founder, Greg Brockman, in federal court, claiming they were putting the commercial interests of the company and A.I. ahead of the public good.
This year, Mr. Musk and a consortium of investors also offered to buy the assets of the nonprofit that controls Open AI for more than $97 billion. OpenAI’s board of directors rejected the bid.
Now the company has notably backtracked from the plan to shift control away from the nonprofit. It is unclear if the new structure, which allows the nonprofit to be OpenAI’s largest shareholder, will affect Mr. Musk’s lawsuit.
A public benefit corporation is often described as an organization designed to create public and social good and allows outsiders to invest in much the same way they invest in other companies.
“I am very happy we made the decision for the nonprofit to maintain control,” Mr. Altman said during a news conference. He added that the new change “sets us up to have a more understandable structure to do the things that a company like ours has to do.”
OpenAI said that it was still negotiating the nonprofit’s stake in the new corporation and that the nonprofit would pick the board members of the new entity.
“I am gratified that the board seems to have worked with regulators and that the nonprofit will maintain control,” said Jill R. Horwitz, a professor of law at Northwestern University who specializes in nonprofits. “But we don’t know what control means yet.”
The Japanese conglomerate SoftBank recently led a $40 billion funding round in Open AI that values the company at $300 billion. If this shift is not completed by the end of the year, SoftBank will have the option to reduce its total contribution to $20 billion, said a person familiar with the latest investment deal.
Mr. Altman said he was confident that the funding would not be reduced.
“We made the decision for the nonprofit to retain control of Open AI after hearing from civic leaders and engaging in constructive dialogue with the offices of the Attorney General of Delaware and the Attorney General of California,” Open AI chairman Bret Taylor said in a statement.
(The New York Times has sued Open AI and its partner, Microsoft, accusing them of copyright infringement regarding news content related to A.I. systems. OpenAI and Microsoft have denied those claims.)