Anthropic safety researcher
Anthropic safety researchers study and test AI systems to reduce risks like harmful outputs or misuse, helping ensure models behave as intended through evaluation and red-teaming.
Standing...
Stories...
Sep 9, 2026
Center
Publicly stated there is a more than 10% chance AI could kill all humans in the next decade.
Lib Dems see this as a wake-up call for urgent, international AI regulation to prevent existential risks, aligning with their evidence-led policy approach.