Former Anthropic safety researcher warns AI poses over 10% chance of human extinction
What happened
A former Anthropic safety researcher warned that rapid AI advancement poses an existential risk, estimating a greater than 10% chance that AI could kill all humans. The warning adds to increasing safety alerts from departing researchers and drew public mockery from Elon Musk, who called it a psy-op.
Why it matters
High-profile safety researcher resignations from frontier AI labs are driving renewed calls from U.S. lawmakers for federal regulation. The public departures intensify political pressure to establish federal rules and safeguards over private companies developing advanced AI models.
Bigger picture
The dispute underscores a broader structural tension between rapid technological advancement and corporate safety governance. Mounting departures are fueling political scrutiny while deepening industry polarization over whether existential risks warrant strict oversight.
Watch next
Monitor U.S. lawmakers for specific federal AI regulation proposals and potential safety governance updates from frontier labs.