Ex-OpenAI Safety Leader Says Sam Altman-Led AI Giant Warned About AI Risks While 'Still Training and Deploying These Dangerous Models'

On Wednesday, former OpenAI safety leader David Robinson said a growing disconnect between the company’s warnings about advanced AI risks and its rapid development pace ultimately contributed to his decision to leave.

Robinson Raises AI Safety Concerns

Robinson, who previously led work on OpenAI’s Safety Systems team, discussed his departure in an interview with Ezra Klein on The New York Times’ "The Ezra Klein Show."

Robinson said he experienced a "fundamental cognitive dissonance" as OpenAI continued developing increasingly capable AI systems despite repeatedly warning about their potential risks.

"Part of what was the fundamental cognitive dissonance for me was we keep publishing these warnings, but ultimately, we’re still training and deploying these dangerous models that we’re warning about," he said.

He also said the company’s internal environment remained highly intense despite public discussions about slowing AI development.

Robinson described the pace as "frenetic" and said employees were "running on fumes."

"It feels almost like a ballet or a kind of dance because all these different functions are kind of all streaming together," he said. "There’s a lot of adrenaline."

Robinson had previously expressed support for slowing the pace of AI development and ultimately left OpenAI, saying he believed he could do more for AI safety outside the company.

Altman Pushes for AI Safety

Last month, OpenAI CEO Sam Altman appeared before the UN Security Council and called for shared AI safety standards while supporting continued development.

Anthropic CEO Dario Amodei also supported stronger AI safeguards and called for slowing development so safety testing could keep pace with increasingly powerful AI systems.

Recent OpenAI Departures

In August, former OpenAI COO Brad Lightcap announced that he was leaving the AI company to "start something new." Lightcap said he had spent most of the previous decade pursuing OpenAI’s mission and believed "mission success" was within sight.

In July, former OpenAI President of Applications Fidji Simo stepped down from her full-time role and transitioned to a part-time advisory position after her recovery from a chronic illness took longer than expected.

Simo said she had taken medical leave following a "severe exacerbation" of the illness.

In March, Caitlin Kalinowski resigned from OpenAI over concerns about potential AI use for warrantless surveillance of Americans and autonomous weapons.

She said these issues were "lines that deserved more deliberation than they got" and described her departure as a matter of "principle, not people."

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.