OpenAI, Anthropic, Google in weeks-long AI safety talks

OpenAI's global policy chief Chris Lehane told reporters Tuesday that the three dominant AI labs — OpenAI, Anthropic, and Google DeepMind — have been coordinating on AI safety for several weeks. The disclosure follows Anthropic CEO Dario Amodei's widely circulated essay published Saturday, which called on the industry to voluntarily slow frontier AI development to reduce catastrophic risks. Sam Altman, Demis Hassabis, and even Elon Musk publicly backed Amodei's call. Altman pledged OpenAI would embed third-party evaluators to monitor safety, mirroring a commitment Anthropic had already made.

The Information reported this week that the three firms are actively working to establish a shared standards body — one Altman reportedly told staff would proceed without U.S. government backing. Hassabis had separately urged Washington in July to create an AI watchdog with authority to screen advanced models and trigger industrywide slowdowns if risks escalated. President Trump has dismissed AI safety concerns as a hoax, while advisor David Sacks called existential risk fears overblown.

Antitrust risk looms over the coordination. Altman himself noted the talks could attract scrutiny if deemed anti-competitive. Amodei's essay proposed a narrow government waiver to legitimize safety-focused coordination, but Lehane said the companies believe no such waiver is needed. Lehane also stated OpenAI supports a FRONTIER Act provision that would require top labs to admit independent verification organizations.