OpenAI Teams Up with Anthropic and Google DeepMind on AI Safety

As concerns grow in the US that AI could threaten the economy and national security, OpenAI has started working with rivals Anthropic and Google DeepMind on AI safety research. OpenAI's global policy chief Chris Lehane said the three companies have been in contact on AI safety for weeks and believe such coordination does not need a special antitrust exemption.

Anthropic CEO Dario Amodei previously published a long essay calling for a slowdown in developing the most advanced AI systems, giving researchers more time to understand potential threats. OpenAI CEO Sam Altman and SpaceXAI head Elon Musk soon expressed support for his stance.

Lehane said it is not unusual for companies to cooperate on safety, pointing to the aviation industry as an example. He added that OpenAI is willing to support bipartisan proposals in the US aimed at preventing potential catastrophic AI risks.