OpenAI, Google, and Meta Sign White House Voluntary AI Safety Agreement

Leaders from OpenAI, Google, and Meta signed a voluntary Superintelligence Safety Agreement at the White House on September 30. Hosted by U.S. President Donald Trump, the agreement establishes a new benchmark for industry self-regulation. Participating companies have committed to rigorous risk assessments for frontier models, external audits by independent third parties, and regular red-teaming exercises to ensure transparency and safety.