HN thread: AI models don't kill people — people do

A Hacker News discussion centers on Anthropic researcher Jacob Coxon's resignation and his views on AI risk; colleague Evan Hubinger echoed concerns, suggesting a >10% chance within a decade. Commenters argue the danger stems from irresponsible corporate deployment rather than the models themselves.

• Insider AI warnings spark Silicon Valley debate2026-09-13
• AI researcher quits Anthropic, warns of deadly risk2026-09-09
• Report: AI already causing deaths and used by multiple militaries2026-09-16
• The right and wrong ways to regulate AI2026-09-16
• Anthropic Signs First Australia Data Centre Lease2026-09-16