AI-Enabled Military Autonomy and Intelligence Failures Put Defense Verification Under Scrutiny

A reportedly fabricated AI-generated intelligence assessment nearly prompted military interception of a Chinese vessel, with additional alleged targeting failures involving stale data and hallucinations. Separately, a Swedish startup reportedly demonstrated autonomous strike drones that selected and attacked targets using onboard AI without communications or human input. These reports make provenance, source checking, meaningful human control, auditability, escalation safeguards, liability, and deployment authorization central questions; the incidents and broader systemic significance remain under investigation.