OpenAI, Google and Anthropic Weigh AI Slowdown as Trump Rejects New Guardrails
Updated
Updated · CNBC · Sep 16
OpenAI, Google and Anthropic Weigh AI Slowdown as Trump Rejects New Guardrails
3 articles · Updated · CNBC · Sep 16
Summary
OpenAI, Anthropic and Google have begun talks on AI safety and whether to slow model development after a researcher’s claim of a more than 10% chance AI could “kill all humans” ignited a social-media storm.
The discussions build on a July proposal from Google DeepMind’s Demis Hassabis for a U.S.-led standards body, as pressure grows on major AI firms to address safety and security risks.
Marc Benioff joined calls for restraint, but Donald Trump attacked any slowdown, arguing strong presidential leadership—not new guardrails—is what AI needs.
Trump’s stance reflects his administration’s view that limits on U.S. AI development would advantage China, setting up a sharper clash between safety advocates and accelerationists.