OpenAI Staff Fund Rival Super PAC Opposing Boss's AI Deregulation Push

A cohort of current and former OpenAI employees are funding an organized political counteroffensive against their company's president. Seven current OpenAI employees and one former employee have donated to Guardrails Alliance, a super PAC that launched last month with $5 million（約8億円） in initial funding and aims to raise $15 million（約24億円） this election cycle to promote stricter AI regulations. The largest contribution came from Juan Felipe Cerón Uribe, a research engineer who has worked at OpenAI since 2022 and donated $200,000. In a statement to WIRED, Cerón Uribe explained that he spent four years developing strategies to mitigate societal harms from AI, but became concerned that "all that research will have gone to waste if it doesn't translate to guardrails that hold private companies accountable." He cited tech billionaires—naming Greg Brockman specifically—as having funded Leading the Future to keep AI unregulated, calling his decision to donate to Guardrails Alliance "easy."

The donations pit employees against their leadership in a high-stakes political battle. Greg Brockman, OpenAI's president and cofounder, and his wife Anna have committed $50 million（約80億円） to Leading the Future, a pro-industry super PAC that has raised more than $100 million（約160億円） in total. Leading the Future's stated goal is to "oppose policies that stifle innovation" and has already spent heavily to sink regulatory efforts—including a $27 million（約43億円） spending blitz in New York's 12th congressional district to defeat Alex Bores, the author of New York's landmark AI safety law, in a primary election last month. OpenAI's global affairs chief, Chris Lehane, helped set up Leading the Future and has consulted with Brockman on his political giving, though the company has tried to distance itself by stating that Brockman's engagement is in a personal capacity. When pressed by employees worried about the company's ties to the super PAC, OpenAI leadership released a June blog post affirming that "employees are free to participate in the political process in their personal capacities."

Several other OpenAI safety researchers are now exercising that freedom in opposition to company leadership. Gabriel Wu, an OpenAI safety researcher, gave $5,000 to Guardrails Alliance and told WIRED he donated "to push back against Leading the Future" and unregulated AI development by the ultra-wealthy. Julie Steele and Jason Wolfe, two OpenAI staffers researching AI alignment, each donated $5,000. David Farhi, a former OpenAI research manager who left after seven years last summer, donated $3,000 and stated in a message to WIRED that after years leading AI research at OpenAI, he became concerned about leading AI research at OpenAI, he became concerned that "AI is going to present our world with both unprecedented opportunities and challenges"—and expressed disappointment that Leading the Future is "actively work[ing] against OpenAI's mission by aiming to shut down" discussion around regulation.

Guardrails Alliance positions itself as a populist counterweight funded by tech workers, labor unions, and other groups rather than major corporate donors. Shaunna Thomas, a longtime Democratic political organizer and Guardrails Alliance cofounder, told WIRED that financial disparity with Leading the Future is not the limiting factor: "We're not going to match our opponents dollar-for-dollar, we don't have to. When you expose what the AI PACs are doing, the people reject it." She noted that after Leading the Future's launch last year, politicians seeking to advance AI regulation faced "a $100 million（約160億円） threat," and that Guardrails Alliance intends to support Democratic candidates in 2026 races, including California's 34th congressional district. A rival effort, Public First Action backed by Anthropic with $20 million（約32億円）, is also building infrastructure to counter Leading the Future in the 2026 elections. Leading the Future's spokesperson, Jesse Hunt, denied accusations of suppressing debate and claimed the super PAC has "advocated for federal regulations on the technology," stating it is "proud of [its] track record supporting a diverse array of policymakers and candidates across the country." The group's public filing is expected to disclose additional donors this week, including former Andreessen Horowitz partner John O'Farrell, who left the venture firm earlier this year and criticized his former colleagues in a New York Times opinion piece for using Leading the Future to "intimidate politicians."

The donations reveal a fault line within OpenAI between leadership's deregulation agenda and rank-and-file employees' concern for AI safety oversight. Greg Brockman's $50 million（約80億円） commitment to Leading the Future—which the body describes as aiming to "oppose policies that stifle innovation"—has prompted internal pushback from workers who spent years developing safety measures. Research engineer Juan Felipe Cerón Uribe's $200,000 donation signals that employees fear their work on AI harm mitigation will be rendered meaningless if companies face no regulatory accountability. This split is not unique to OpenAI: Anthropic has already committed $20 million（約32億円） to Public First Action, another super PAC promoting AI safeguards, suggesting the entire AI industry is divided on whether stronger government oversight serves or harms the sector's long-term interests. Shaunna Thomas, Guardrails Alliance's cofounder, argues that the financial gap between her group ($15 million（約24億円） goal) and Leading the Future ($100+ million) is less important than leveraging public opinion—a bet that voters and politicians will reject the AI industry's unregulated vision once its lobbying is exposed.