Anthropic AI Safety Lead Mrinak Sharma Resigns

An executive resignation at leading AI firm Anthropic has gone viral after Mrinak Sharma, head of the company’s Safeguards Research Team, announced his departure and shared a detailed letter explaining his decision. Posting on social media, Sharma stated that it was his final day at Anthropic and warned that “the world is in peril”, citing a range of interconnected global crises.

In his letter to colleagues, Sharma reflected on the growing risks posed by artificial intelligence, bioweapons, and broader systemic challenges. A UK-based researcher and Oxford PhD graduate, Sharma spent two years at Anthropic working on AI safety, including research on AI sycophancy, bioterrorism safeguards, and the psychological impact of AI chatbots on users. He emphasized that humanity’s wisdom must grow alongside its technological power to avoid serious consequences.

Sharma clarified that he has no immediate next role planned and will return to the UK for the time being. His exit comes amid several recent departures from Anthropic, though reports suggest these moves are independent and do not indicate a mass exodus. Still, the resignation has reignited discussions around AI safety, ethics, and the pressures faced by researchers in the rapidly evolving AI industry.