Leading AI researcher quits Anthropic, cites 'perilous' world
Technology
Mrinank Sharma, a leading AI researcher with degrees from Oxford and Cambridge, has stepped down from his role at Anthropic AI.
He led the Safeguards Research Team and announced his resignation on X (formerly Twitter), where his post quickly gained over a million views.
Sharma's letter highlights ethical concerns in AI development
In his resignation letter, Sharma warned that "the world is in peril" due to overlapping threats—not just from AI, but also bioweapons or AI-assisted bioterrorism.
He shared concerns about pressure to compromise core values at Anthropic and said he's now shifting focus to poetry and "courageous speech," hoping to stay true to what matters most to him.