OpenAI calls Stuart Russell a "doomer" in court after its CEO co-signed his AI extinction warning

OpenAI is trying to silence an AI safety expert in court, using arguments that directly contradict the company's own history of raising the same concerns.

Civil society organization The Midas Project is calling out OpenAI's double standard in the ongoing court battle with Elon Musk. OpenAI filed a motion to exclude testimony from AI professor Stuart Russell, calling him a "prominent AI doomer" who has "made a career giving public lectures warning that AI might kill off humanity." The filing dismisses his views as "dystopian," "speculative," and "alarmist." Russell has spent years warning that advanced AI systems could pose an existential threat to humanity if not properly controlled.

The hypocrisy is hard to miss. OpenAI leaned on exactly these kinds of threat scenarios to build its brand and push its own agenda, from warnings about "scary" mass job losses and social manipulation to outright doomsday scenarios.

The Midas Project also points out that OpenAI CEO Sam Altman co-signed a declaration with Russell in 2023 stating that the risk of extinction from AI should be a global priority. Altman said back in 2015 that AI would "probably, most likely, sort of lead to the end of the world."

OpenAI's lawyers also argue that concerns about an AI race at the expense of safety aren't relevant to the case. But OpenAI's own charter, as The Midas Project notes, spells out exactly this fear—that the late stages of AGI development could turn into a race without enough safety precautions.

OpenAI already seems to be living out that pattern. Safety testing for GPT-4o was reportedly rushed under time pressure, and newer models have been affected too. The company also cut jobs on its safety teams, and several high-profile researchers walked out the door.

OpenAI and The Midas Project have a history

The Midas Project says OpenAI is making opportunistic arguments in court that contradict its own long-held positions. The two organizations already have a track record: OpenAI previously hit The Midas Project with a sweeping subpoena as part of the Musk dispute.

According to founder Tyler Johnston, the company demanded all communications with journalists, lawmakers, former OpenAI employees, and other civil society groups, even though The Midas Project wasn't even a party to the Musk lawsuit.

At the time, Johnston suspected the move was a targeted play timed to coincide with decisions by California and Delaware authorities on OpenAI's conversion to a for-profit company with a valuation of up to $500 billion. The organization had publicly criticized this restructuring.