Hosts: Lenar Kess, Damra Vol. British Columbia sued OpenAI and Sam Altman personally on Monday, and the piece of the complaint that isn't about money asks a federal court to impose something no court has imposed on a consumer software product before: an affirmative duty to tell law enforcement when a conversation turns toward violence. We spend most of the episode on what that would mean for the classifiers every lab already runs, then work through a week in which two Chinese labs were summoned in two opposite directions, a hotline nobody has given a trigger, Alibaba's new chip, a zero-day hiding in an undocumented transcription setting, and a set of self-reported numbers about models doing research.British Columbia's complaint against OpenAI and Sam Altman, filed in federal court in San FranciscoThe Information on the Cyberspace Administration of China opening an investigation into DeepSeek and MoonshotLaurie Chen for Reuters on the UN Security Council inviting Chinese AI companies to brief itFinancial Times on researchers at the UK's AI Safety Institute signing off with stressOpenAI's standards proposal, routed through the Commerce Department's Center for AI Security and InnovationAlibaba's Zhenwu V900 announcement and Eddie Wu's remarks on a five-to-ten-trillion-parameter modelPatrick Wardle's write-up of the Muse zero-day and the undocumented transcription settingTimnit Gebru and Emily Bender in MIT Technology ReviewThe Hacker News report of an agent signing a contract, from a single developer accountIntrinsic Core released under Apache 2.0Isaac ROS 5.0The a16z academy, with Anthropic and Meta contributing compute
