On September 30, 2026, executives from Google, Anthropic, Meta, OpenAI, xAI, and Nvidia signed "The White House Accord on Super Intelligence" during a luncheon hosted by President Donald Trump. The accord is described by Trump as an act of "tremendous self-regulation," but it does not represent the comprehensive AI regulation that many within the companies have been advocating for. The agreement outlines several voluntary commitments, including the implementation of robust internal controls to monitor AI models, the establishment of internal teams for oversight, partnerships with external monitors for independent assessments, and the formation of board committees to oversee these efforts.
While the accord indicates an effort by these companies to enhance AI safety, hopes for substantial government endorsement were diminished by reports that the Federal Trade Commission (FTC) plans to investigate several AI companies for potential consumer protection violations. This accord follows previous voluntary agreements on AI safety practices, including commitments made by UK and South Korean companies in early 2025.
Neil Chilson, a former chief technologist for the FTC, noted on social media that the FTC could enforce the commitments if companies fail to adhere to them. However, skepticism remains regarding the potential for significant regulatory action, especially in light of the recent luncheon with Trump, which some believe may influence the FTC's approach to enforcement. The FTC has confirmed an investigation is underway but has not provided details on the specific consumer protection issues being examined.
The FTC's chairman, Andrew Ferguson, has indicated that investigations often begin without a fully developed case. Douglas Farrar, a former FTC public affairs head, expressed concern that political dynamics may hinder any serious legal action against the AI firms involved in the accord. The situation continues to evolve as the FTC navigates its investigation and the implications of the recent accord.