OpenAI Proposes Global AI Safety Standards Ahead of Sam Altman's UN Security Council Briefing

OpenAI has proposed international standards for measuring frontier-AI capabilities, classifying incidents and reporting model misalignment, days before Sam Altman is expected to brief the UN Security Council on AI safety on Wednesday, September 23, 2026. The move links a new disclosure framework and six incident reports to a broader push for shared rules among labs, governments and AI safety institutes.