OpenAI, Anthropic and Google DeepMind have been discussing AI safety for weeks, policy chief says
OpenAI confirmed that it has been holding safety talks with Anthropic and Google DeepMind for several weeks, while political figures debate regulatory approaches.
OpenAI’s global policy chief, Chris Lehane, told reporters on Tuesday that the company has been in ongoing discussions with rivals Anthropic and Google DeepMind about AI safety, confirming earlier reports. Lehane, who is in Washington to meet with lawmakers, said the talks have been underway for several weeks and involve coordination on safety standards without requiring a government waiver.
The disclosure follows an essay by Anthropic CEO Dario Amodei that urged the industry to collaborate on slowing the development of frontier AI to avoid catastrophic risks. The letter was endorsed by several AI leaders, including OpenAI CEO Sam Altman, who said OpenAI would work with Anthropic to embed third-party evaluators for safety monitoring. Google DeepMind’s chief scientist Demis Hassabis has also called for a new standards body to act as an industry watchdog.
Altman has previously hinted at private discussions with other AI executives and indicated that a standards body could be created without direct U.S. government involvement. Some observers have warned that such coordination might raise antitrust concerns if it is seen as suppressing competition.
In parallel, President Donald Trump dismissed AI safety concerns as a hoax and argued that regulatory slowdowns could advantage China. His AI adviser David Sacks echoed the view, describing existential risk fears as exaggerated. At the same meeting, Lehane said OpenAI backs a provision in the proposed FRONTIER Act that would require leading AI labs to permit independent verification organizations to assess model safety.