[AINews] not much happened today
US may ban Chinese open-weight AI models as Hugging Face discloses using open models for cyber defense
The Trump administration is reportedly considering a de facto ban on cutting-edge Chinese open-weight models like Kimi through procurement restrictions, Entity List designations, and liability regimes rather than a clean statutory ban. Simultaneously, Hugging Face disclosed that during a cyber incident they relied on self-hosted open-weight GLM-5.2 for forensic analysis because commercial frontier API guardrails blocked the work and sensitive attacker data had to stay on-prem. Clement Delangue and other prominent technical voices argued that restricting open models would harm competition, national sovereignty, and defensive security more than it would help US incumbents.