Policy & Safety
Top in Policy & Safety
Microsoft agrees to binding AI limits for US schools from November 1
3 min read16 Sept 2026
The EU AI Act, regulation, court cases and AI safety — what the rules mean in practice.
Top in Policy & Safety
EU government officials gather in Brussels on Thursday to discuss global rules for new AI models, with the European Commission due to give a technical briefing on recent AI incidents and updates on international AI policy, according to two EU diplomats and a draft agenda seen by Politico Europe.
The Manhattan District Attorney's Office has seized 12 domains it says were used to unlawfully share, publish and sell celebrity deepfake videos. Around 1,200 people, overwhelmingly women, were depicted in nonconsensual sexual images and videos on the sites, the office says.
Jacob Tsimerman, a 2026 Fields Medalist, is scientific director of the new Mathematical AI Safety Institute, which aims to prove AI systems safe the way cryptographers prove codes unbreakable.
OpenAI has asked Congress whether coordinating a slowdown in AI development with rival labs would violate US antitrust law, after chief scientist Jakub Pachocki called for shared safety standards.