The Trump administration has opted against creating a new federal regulatory framework for artificial intelligence. Instead, the government plans to utilize existing liability laws to address potential risks associated with the technology. This approach is supported by Treasury Secretary Scott Bessent and former White House AI czar David Sacks.
This strategy follows a White House meeting held on September 29-30, 2026, where President Trump endorsed a non-binding agreement with major AI companies, including Anthropic, OpenAI, Nvidia, Meta, and Google. The accord emphasizes internal company processes and independent audits rather than legally enforced standards.
The debate over regulation remains complex. A March 2026 report from RAND, which surveyed 24 experts, indicated that strict regulations are considered the least feasible option. Instead, experts identified incentives for risk disclosure and voluntary measures as more realistic paths for managing AI development.
Dissent exists within the industry. On October 2, 2026, Google executive James Manyika stated that self-policing is insufficient for critical AI technology. Meanwhile, California Governor Gavin Newsom issued an executive order on September 18, 2026, to review AI safety plans, including potential requirements for emergency shutdown mechanisms.
