[OpenAI, Google, and Meta Sign Voluntary Agreement at the White House, Commit to External AI Audits] On September 30, company leaders from OpenAI, Google, and Meta signed the 'Superintelligence Safety Agreement' at the White House, with U.S. President Trump presiding over the meeting. Participating companies agreed to conduct risk assessments on frontier models, accept external audits carried out by independent third-party organizations, and commit to regular red team testing, aligning certain safety benchmark data with regulatory frameworks. (Source: CoinDesk)