OpenAI calls for mandatory national AI safety rules

OpenAI has called on Congress to enact mandatory, capability-based national AI safety regulation before it adjourns in December, marking a significant shift by one of the world's leading artificial intelligence companies toward binding federal rules. The proposal was set out in a blog post published Wednesday by Chris Lehane, the company's chief global affairs officer, who argued that recent jumps in AI capabilities demand more than voluntary commitments. The company urged requirements including common testing standards, independent assessments, stronger cybersecurity protections, clear incident-reporting rules, mandatory monitoring for model misalignment, prompt written notice when a model circumvents security controls, and alignment-evaluation gates before deployment. OpenAI also said it supports voluntary industry standards and international coordination on how advanced AI capabilities are measured, risks are managed, and human control is maintained, while arguing that compatible international standards will be necessary for the United States to lead globally.