AI guardrails face live battlefield tests as Houthis expand strikes
Anthropic’s 154-page report catalogs weaponization, surveillance, and influence attempts—including a Yemen-based cell seeking missile guidance software—while Houthi attacks hit Saudi military and energy sites with implications for shipping and regional alignments.
Anthropic released a 154‑page threat report detailing attempts to misuse its Claude AI—from surveillance and influence operations to weapons development—including a Yemen‑based cell in Houthi‑held ...