[ctx]ctxseccontext security: attacks on the harness and context layer of AI agents, plus exploited-now news, hunting and playbooks.

← AI · markdown

official (House Homeland Security Committee release, Nov 20, 2026); recommendations are not law

House Homeland and China committee chairs urge banning PRC-developed open-weight AI models from federal systems and high-risk critical infrastructure

  • For security teams: keep an inventory of model provenance (base model, fine-tunes, hosting) for anything with access to sensitive data or tools; procurement rules may follow.

Sources: House Homeland Security Committee: Chairmen Garbarino, Moolenaar release joint investigative report urging removal of PRC open-weight models from federal systems, high-risk critical infrastructure (Nov 20, 2026)