OpenAI Staffers Are Funding a Rival Super PAC to Take on Their Boss

Something odd is happening inside OpenAI's ranks: employees are putting their own money into a political fight against their boss. At least seven current staffers and one former researcher have donated over $215,000 to Guardrails Alliance, a super PAC pushing for tougher rules on frontier AI labs. The group launched last month with $5 million in seed funding and wants to raise $15 million total this cycle, aiming squarely at Leading the Future, the pro-industry PAC that Brockman has committed $50 million to alongside his wife Anna.

The biggest single donation, $200,000, came from Juan Felipe Cerón Uribe, a research engineer who has spent four years at OpenAI working on ways to reduce AI's societal harms. He says he doesn't want that work wasted if it never turns into actual accountability for companies building the technology. Gabriel Wu, a safety researcher, gave $5,000 and put it plainly: he's worried about a handful of ultra-wealthy people deciding AI's future without any check on their power. Alignment researchers Julie Steele and Jason Wolfe each chipped in $5,000 too, and former research manager David Farhi, who left last summer after seven years, donated $3,000 and said Leading the Future is working against the very mission OpenAI claims to have.

The money is small next to Brockman's commitment, and Guardrails Alliance cofounder Shaunna Thomas isn't pretending otherwise. She argues her group doesn't need to match Leading the Future dollar for dollar because public opinion already leans toward regulation — it's cheaper to tap into sentiment that exists than to buy it. Her group already has a scoreboard entry: it backed Alex Bores, the New York assemblyman behind the state's AI safety law, who lost a primary last month after Leading the Future spent heavily against him. That race alone saw $27 million in combined PAC spending, with Anthropic-backed Public First Action also weighing in on Bores' side.

OpenAI's official line, via a June blog post, is that Brockman's involvement with Leading the Future is personal, not corporate, and that employees are free to donate to whatever causes they choose. That's technically true, but it glosses over the obvious tension: the company's president is bankrolling an effort to block the kind of oversight some of his own researchers spent years building a case for. Leading the Future spokesperson Jesse Hunt pushed back on claims the group tries to suppress debate, pointing to its past support for federal AI rules.

Guardrails Alliance plans to disclose more donors this week, including former Andreessen Horowitz partner John O'Farrell, who left the firm this year and later wrote publicly that Leading the Future was being used to intimidate lawmakers who ask hard questions about AI governance. Whether a scrappy coalition of tech workers and unions can meaningfully offset $100 million in industry money is unclear. But the fact that OpenAI's own employees feel compelled to organize against their employer's political spending says plenty about how unsettled the internal debate over AI regulation still is.