AI Governance
Data Centers
Drones
Climate
Privacy
2026 Maine Senate
Bob Dylan
Vance Pushes Human Control Over Military AI
Coverage from NBC Chicago, NBC News, and others
Vice President JD Vance is emphasizing that humans must retain authority over life-and-death decisions as artificial intelligence becomes more capable on the battlefield.
His position comes amid a dispute between the Trump administration and Anthropic over military-use guardrails, as well as White House debate over how closely to monitor frontier models after Anthropic reported that Mythos Preview identified thousands of serious software vulnerabilities. A proposed executive order for voluntary government testing of leading AI models was reportedly not signed, leaving the administration’s approach unsettled.
The main update is that the internal policy debate has become more concrete: Trump reportedly refused to sign a proposed voluntary AI-testing executive order, leaving the administration without a settled oversight path. At the same time, the cybersecurity rationale for tighter monitoring has been sharpened by Anthropic’s reported vulnerability-finding results.
The main update is that the policy debate has become more concrete: a drafted executive order for voluntary AI model testing existed but was never signed, and White House cyber/policy staff are now more directly involved. The cybersecurity side is also sharpened by a specific Anthropic model, Mythos Preview, being tied to severe vulnerability findings.