Azernews.Az

Akbar Novruz

Read more

Mrinank Sharma, the artificial intelligence safety lead at Anthropic, has resigned from the company, issuing a stark public warning that “the world is in peril” amid a convergence of global risks, AzerNEWS reports.

In a letter published on Monday, Sharma said he had struggled to ensure that his actions were consistently guided by his core values, suggesting that he faced internal pressures to set aside what mattered most, though he did not provide specific details.

Sharma joined Anthropic in 2023 after completing a PhD at the University of Oxford. He led the company’s Safeguards Research Team, focusing on high-risk areas of AI safety, including preventing AI-assisted bioterrorism and examining how advanced chatbots can distort users’ perception of reality.

Miles Deutscher, a widely recognized prominent figure in the tech and startup ecosystem, stated that the resignation comes at an interesting period.

In his farewell note, Sharma warned that humanity is approaching a critical moment in which its growing technological power is outpacing its collective wisdom. Without a corresponding increase in responsibility and ethical restraint, he argued, the consequences could be severe.

He said he plans to return to the United Kingdom, withdraw from public life “to become invisible,” and devote himself to writing poetry.

Sharma’s departure adds to mounting concerns within the tech industry and beyond about whether AI developers can truly prioritize safety, ethics and long-term human values in an environment increasingly driven by speed, competition and commercial pressure.