OpenAI Greg Brockman AI safety PAC

작성: Tim Official: Greg Brockman

OpenAI president Greg Brockman has been exposed for backing a fake AI safety movement intended to undermine genuine efforts. Taylor Lorenz and Tyler Johnston detailed the scheme in their investigation, and the findings reveal a coordinated effort to manipulate both sides of the debate and generate artificial outrage. In July 2025, exactly 10 days after a dark money offshoot of the pro-AI super PAC Leading the Future filed its articles of incorporation in Nevada, a new X account named Doomers Are Dumb surfaced. At first it presented itself as a typical meme page, sharing jokes about waifus and dating issues. Then in October, just as the PAC started its public campaign, the account shifted to posting lewd and offensive material that targeted AI safety advocates. On the same day, another account called Jonathan Doomer emerged on X. It portrayed a frantic AI critic who had left his job to issue alarms about the risks of the technology. This figure served as an ideal strawman, appearing loud, erratic, and simple to criticize, while the Doomers Are Dumb account engaged it from the outset. The PAC effectively controlled both positions in the discussion. It created the phony critic to tarnish legitimate skeptics by linking them to an unstable counterpart. The financial backers of this network are far from obscure. Greg Brockman contributed 25 million dollars to Leading the Future and another 25 million dollars to Trump's MAGA Inc. Andreessen Horowitz, Joe Lonsdale of Palantir, Ron Conway, and Perplexity each provided substantial funding. Overall pledges reach approximately 140 million dollars, with around 51 million dollars yet to be delivered. The political strategist overseeing the entire effort is Josh Vlasto, who previously directed crypto's 300 million dollar Fairshake PAC that influenced the outcome of the 2024 election. Their focus is on the 38 state-level AI safety laws that legislatures approved in 2025. The first major target is Alex Bores, who is running for Congress in New York's 12th district. Bores had worked at Palantir until 2019, when he resigned in protest over the company's contracts with ICE. Afterward he authored New York's initial state-level AI safety legislation. The PAC has pledged at least 10 million dollars to oppose him, and its advertisements accuse him of building the tech for ICE deportations, the very activity he had abandoned. Yet the most troubling development came from OpenAI yesterday. The company released a statement attempting to separate itself, noting that its employees remain free to engage in the political process on their own. In effect, it suggested the actions belonged to its president and executives rather than the organization itself. Then Jason Kwon, OpenAI's chief strategy officer, was discovered following the fake accounts on X. These profiles have almost no followers, so they are not the sort one stumbles upon by chance. One follows them only with prior knowledge of their existence. Chris Lehane, OpenAI's chief global affairs officer, participated in the initial discussions that established the PAC, according to the Wall Street Journal. The firm that has applied for an 852 billion dollar IPO continues to tell Congress it backs thoughtful AI regulation. At the same time, its leaders are financing a 140 million dollar campaign aimed at removing every lawmaker who seeks to implement such rules. This marks the second controversy for the PAC this year. Back in May the same group was found paying TikTok and Instagram creators 5,000 dollars for each post to circulate anti-China AI panic material prepared by political operatives. This pattern shows what the AI billionaire class pursues while its public relations efforts place Sam Altman on podcasts discussing a desire for oversight. The approach involves supporting fabricated opponents to make authentic ones seem irrational, compensating creators to generate manufactured crises when needed, and directing every aspect of the dialogue to prevent any meaningful discussion of AI safety from taking place.