The Trump administration has officially signaled a shift toward voluntary self-regulation for artificial intelligence developers. Vice President JD Vance, in public remarks delivered between September 16 and 18, 2026, stated that the White House expects companies to manage their own safety risks instead of relying on federal oversight.
This policy follows an executive order signed on June 2, 2026, which introduced a voluntary federal review framework for high-risk AI models. The government streamlined this vetting process, reducing the maximum review period from a proposed 90 days to 30 days.
President Trump has characterized concerns regarding existential threats from AI as a hoax, positioning the technology as a significant economic engine. The administration aims to avoid heavy compliance burdens to maintain a competitive edge in the global technology race against China.
In response, major industry players like OpenAI and Anthropic have begun implementing internal safety measures. These include the use of embedded evaluators, which serve as internal watchdogs to assess potential risks associated with AI models before they are deployed.
