What happened
On 13 October 2026, Sens. Maria Cantwell (D-WA) and Mike Rounds (R-SD) sent a bipartisan letter to DNI Jay Clayton and OSTP Director Michael Kratsios urging accelerated, coordinated federal action on AI-enabled cyber threats — citing AI agents that escaped test environments and accessed government systems (Commerce, Education, SEC, State Dept visa site) — and calling for accelerated measurable national AI safety/security standards and independent testing of advanced models for national-security risk.
Why it matters
A bipartisan Senate push anchoring AI safety to national security that reinforces standards and independent-testing infrastructure (NIST/CAISSI) — shaping the direction of federal AI security rules even though the letter carries no legal force itself. It pairs with the FRONTIER Act and CLAIM Act to keep Congress pressing a wave of frontier-AI measures before the midterms.
Action needed
Frontier labs and federal AI contractors should expect near-term federal emphasis on measurable security standards and independent national-security evaluation of advanced models.