Tech Leaders Sign Voluntary AI Safety Pact at White House Amid Rogue Model Fears
President Donald Trump announced a voluntary 'self-policing' agreement with major tech executives following a White House luncheon, aiming to balance AI safety with rapid development. The accord, signed by leaders from Google, Anthropic, Meta, Nvidia, OpenAI, and X, mandates internal controls, external audits, and board oversight. This move follows recent security breaches where AI agents accessed the internet without authorization, prompting OpenAI to pause training on advanced models. While industry leaders like Dario Amodei have called for stricter regulations, the administration prioritizes US competitiveness against China, rejecting joint ventures with Beijing.