Some OpenAI employees have started quietly bankrolling a campaign against their own president. More than $215,000 of their own money has gone to Leading the Future’s chief rival—a super PAC that wants exactly the kind of AI rules their boss is spending millions to prevent. That boss is Greg Brockman, OpenAI’s president and co-founder, who helped seed Leading the Future with a chunk of its $100 million-plus war chest. The staffers are backing Guardrails Alliance, a group launched last month with $5 million to push for tougher oversight of frontier AI labs. First reported by Wired, the split lays bare a widening rift inside the company over who should write the rules for the technology it’s racing to build.
One engineer put up $200,000 of the total
Most of that money came from a single person. Juan Felipe Cerón Uribe, a research engineer at OpenAI since 2022, gave $200,000 on his own. His job for the past four years has been reducing the harms AI might cause society—and he’s come to believe that work is wasted without rules to back it up.“Tech billionaires, such as Greg Brockman, funded the super PAC Leading the Future to keep AI unregulated,” Cerón Uribe told Wired, calling his decision to donate an easy one. The rest of the staffers gave far less but sang the same tune. Safety researcher Gabriel Wu put in $5,000 to push back against the flood of money aimed at keeping AI unregulated. Alignment researchers Julie Steele and Jason Wolfe each added $5,000.
The money gap is huge, and that’s the plan
Next to Brockman’s numbers, the donations look small. He and his wife Anna have pledged $50 million to Leading the Future—more than triple what Guardrails Alliance hopes to raise all cycle. The employee checks are a sliver of that $15 million goal.Shaunna Thomas, who co-founded Guardrails Alliance, isn’t losing sleep over the mismatch. The longtime Democratic organizer told Wired she doesn’t need to match her opponents dollar for dollar. Her bet is that voters turn on these AI PACs once they see what the spending is actually for. Exposing the money, she argues, costs a lot less than outspending anyone.That approach already has a track record. Leading the Future’s opening move was an effort to sink Alex Bores, author of New York’s landmark AI safety law. Bores lost his primary last month, in a race that pulled in a staggering $27 million from pro-industry and pro-safeguard groups combined.
OpenAI is trying to back away
The donations put OpenAI in an awkward spot. Brockman’s ties to Leading the Future have unsettled enough employees that they’ve pushed leadership for answers, and the company has since worked to distance itself from the group.Asked for comment, an OpenAI spokesperson pointed to a June blog post calling Brockman’s involvement personal, not corporate—and noting that employees are free to back campaigns and causes on their own time. Chris Lehane, OpenAI’s global affairs chief, has said he helped set up Leading the Future and advised Brockman on his giving, though he isn’t part of daily operations. The PAC rejects the charge that it muzzles debate; spokesperson Jesse Hunt said it has pushed for federal AI rules and stands by its record of backing a range of candidates.Guardrails Alliance isn’t fighting alone. Public First Action, a super PAC seeded with $20 million from Anthropic, is working the same side of the field this cycle. Both groups backed Bores. Thomas says she’s now hunting for other Democrats to support in 2026, starting with a candidate in California’s 34th district. From here, the filings—and the fault lines—only get easier to see.
