Particle.news
Download on the App Store

Vance: Humans Must Keep Final Say Over Lethal AI Decisions

His Air Force Academy remarks highlight rising concern inside the administration over cybersecurity flaws in AI models, raising questions about the future of military guardrails.

Overview

  • Vice President JD Vance told Air Force Academy graduates on Thursday that decisions over life and death on the battlefield must remain human judgments and urged officers to use AI only as a tool, not a replacement for moral choice.
  • Vance referenced Pope Leo XIV’s recent theological document on AI in his speech and said his main worry about AI is how it will transform warfare and the conduct of combat.
  • The comments follow administration alarm over Anthropic’s Mythos Preview, which reported thousands of severe software vulnerabilities and prompted urgent White House discussions with tech CEOs about model security and testing.
  • The administration remains divided on policy: some officials pushed a draft executive order to create voluntary government testing of advanced models but President Trump did not sign it, and Defense Secretary Pete Hegseth has clashed with Anthropic over lifting company guardrails for military use.
  • The debate now centers on whether to tighten cybersecurity oversight and preserve strict limits on military applications of AI, a choice that could shape procurement, battlefield rules, and how officers are trained to integrate — but not defer to — automated tools.