The Trump administration chose in early October 2026 to rely on existing liability law rather than create a new regulatory framework specifically for artificial intelligence. This approach is backed by Treasury Secretary Scott Bessent and former White House AI czar David Sacks, and follows a non-binding agreement signed in September with major AI companies including Anthropic, OpenAI, Nvidia, Meta and Google. A RAND report published in March, however, shows that most experts do not believe strict regulation is feasible. Dissent has also emerged from within the industry: Google’s James Manyika stated on October 2 that self-regulation was no longer sufficient. California has simultaneously launched its own review of AI safety, including mechanisms to shut down AI systems if they behave dangerously.
Source: Read the original article
