WorkSeptember 9, 2026via CNBC Technology
Anthropic researcher says AI has more than 10% chance of 'killing all humans' after colleague quits
Why it matters
High-profile safety departure and public risk statement from inside a frontier lab signals internal disagreement on risk thresholds and governance — a workforce/culture story that shapes how practitioners think about vendor trustworthiness and how the field approaches existential safety.
Key signals
- Anthropic researcher stated >10% probability of AI killing all humans
- Colleague departed citing safety concerns
- Published September 9, 2026
- Signals internal disagreement at frontier lab on risk tolerance
- Public statement from safety researcher — not private memo
- Anthropic safety researcher departure over safety concerns
- Colleague's quantified existential risk estimate: >10% probability of human extinction from AI
- Signals internal safety-vs-capability tension at major frontier lab
The hook
Anthropic safety researcher pegs existential AI risk at >10% as colleague exits over safety concerns.
An Anthropic safety researcher said there is a greater than 10% chance AI could "kill all humans" after a former colleague quits over safety concerns.