OpenAI just cut a deal with California. Critics declare it’s full of holes

OpenAI just cut a deal with California. Critics say it’s full of holes


OpenAI declared Tuesday it would restructure as a for-profit company in a way that addresses concerns from California Attorney General Rob Bonta, who signed off on the transformation.

But details of the shift could revive worries that OpenAI is misapplying charitable tax exemptions, experts and advocates informed CalMatters. The ChatGPT buildr is putting its nonprofit arm nominally in control of the for-profit entity, but there are numerous ways the for-profit company could finish up calling the shots, these people declared. There are also important, unanswered questions about the safeguards that are supposed to keep that from happening.

Under the restructuring, the newly-formed OpenAI Foundation will hold about 26 percent of OpenAI’s valuation, a share amounting to $130 billion, instantly building it one of the most well-finidisplayed philanthropic organizations in the world. Microsoft, company employees, and other investors will hold the rest. The controlling nonprofit foundation can appoint members of the for-profit board of directors and, through a special committee, step in to address AI safety concerns. The company also pledged to remain in California.

OpenAI did not respond to a CalMatters request for additional details about potential safeguards to preserve the indepfinishence of the OpenAI Foundation.

OpenAI’s plans came under scrutiny in California becautilize Bonta, along with Delaware Attorney General Kathy Jennings, wanted to ensure the company stayed true to the mission laid out in its charter when the organization was founded as a nonprofit a decade ago to build artificial innotifyigence that benefits humanity. The company had pledged all “assets are irrevocably dedicated” to this purpose.

OpenAI has faced criticism for a wide range of impacts on society. In August, the parents of California teenager Adam Raine alleged in a lawsuit that ChatGPT coached him on how to commit suicide. The company put restrictions on its generative AI video app Sora 2 after depictions of Martin Luther King Jr were criticized as disrespectful. Lawbuildrs in California have also shiftd to mitigate rising power consumption and proliferation of data centers driven by ChatGPT and similar tools. At the same time, the company has supported drive an AI boom that has seen Big Tech companies surge money into state tax coffers.

Bonta and Jennings have both now signed agreements with OpenAI blessing its new structure.

“We will be keeping a close eye on OpenAI to ensure ongoing adherence to its charitable mission and the protection of the safety of all Californians,” Bonta wrote.

Robert Bartlett, a professor of law and business at Stanford Law School, has studied and worked in the venture capital ecosystem for three decades. He declared OpenAI’s start as a nonprofit was unusual and related to its unique mission around artificial innotifyigence. But it found being a nonprofit restrictive, building it difficult to raise capital and compensate its employees with equity in the company. Its restructuring should pave the way for an eventual initial public offering.

Bartlett declared the new arrangement that the nonprofit, a minority stakeholder, will have oversight of the public benefit corporation is also unusual. He declared the deal envisions a “pretty active role” for the nonprofit’s safety committee, which will include the right to control safety procedures and halt the release of AI models created by the corporation. OpenAI previously named four members of the safety committee on its website and has declared all current members of the non-profit board will serve on the for-profit board, with some as observers.

But not knowing exactly how much overlap there might be between the boards of the nonprofit and the corporation is a huge question, as is the ultimate composition of the committee, Bartlett declared.

“We’ll have to see what happens, who’s on the committee, how active (they are), and their relationship to OpenAI,” Bartlett declared. “Will (the structure) be meaningful and consistent with the AG’s focus on safety?”

Steven Adler previously led a product safety team at OpenAI. On Tuesday he published an op-ed in the New York Times that argues that the company can’t be trusted when they declare they can safely deploy erotica chatbots in part becautilize it has a history of ignoring risks.

He informed CalMatters that under the restructure that he considers the nonprofit’s safety committee necessarys more indepfinishence to operate effectively. “I hope that a truly indepfinishent body will do a better job of protecting the organization’s mission than one that feels any pull toward profits,” he declared.

There’s a bazillion conflicts of interest here.

— Judith Bell, San Francisco Foundation

OpenAI’s restructuring drew ire from Eyes On OpenAI, a coalition of more than 60 California nonprofit organizations who have argued for more than a year that attorneys general should force the company to transfer its assets to an indepfinishent nonprofit entity. The precedent for this approach comes from Blue Cross of California, which started as a nonprofit. Following a transfer of assets to a for-profit subsidiary in the 1990s, that organization gave more than $3 billion in stock to two foundations.

San Francisco Foundation chief Impact Officer Judith Bell, a member of the Eyes on OpenAI coalition, declared the deal could set a precedent for startups to evade taxes, and is also concerned that under the restructuring the same people can serve on boards of directors for the for-profit and the nonprofit.

“There’s a bazillion conflicts of interest here,” she declared, adding that those conflicts are particularly worrisome given the broad potential harms the foundation necessarys to keep an eye on, including how the tech impacts children, the economy, the workplace, and society.

The deal speaks to the tremfinishous influence of a corporation to push forward a deal, declared Orson Aguilar, director of the advocacy nonprofit LatinoProsperity and a member of the Eyes On OpenAI coalition.

He believes OpenAI lost its way when key executives realized they could build an enormous amount of money for themselves. Members of the nonprofit board, meanwhile, variously quit and lost influence after some of them attempted to oust CEO Sam Altman in 2023.

“The nonprofit continues to operate under the influence of the for-profit it supposedly oversees and that’s been our hugegest objection and nothing today notifys us that anything meaningful has alterd that,” he declared.

This article was originally published on CalMatters and was republished under the Creative Commons Attribution-NonCommercial-NoDerivatives license.





Source link