Mrinak Sharma, AI safety lead at Anthropic, resigns
Technology
Mrinak Sharma, who led AI safety at Anthropic, has resigned, sharing worries about "The world is in peril." global threats and the need for wisdom to keep up with fast-moving tech.
He hasn't shared what's next but is heading back to the UK.
Sharma's role and recent departures
During his time at Anthropic, Sharma tackled issues like AI sycophancy and set up guardrails against bioterrorism risks from AI.
His exit follows other recent departures from the company—Harsh Mehta, Behnam Neyshabur, and Dylan Scandinaro (now at OpenAI)—though these moves aren't part of any bigger trend.