Top News

‘The world is in danger, everything is falling apart…’ AI safety engineer Mrinank Sharma made a shocking revelation and resigned.
Samira Vishwas | February 11, 2026 3:24 AM CST

New Delhi: Anthropic Company, known for its cloud-based AI model Cloud, is one of the fastest growing companies in the world. The company’s goal is to create safe and useful AI. But on February 9, 2026, Mrinank Sharma, head of the company’s Safeguards Research Team, suddenly resigned. He shared his resignation letter on X, in which he talked about the world crisis and moral dilemma.

Mrinank’s journey

Mrinak Sharma has obtained a PhD in Machine Learning from Oxford University and a Master of Engineering degree from Cambridge University. He was working at Anthropic for two years. He worked on important projects like AI psychofancy, defense system to reduce the risk of bioterrorism through AI and AI safety case.

Main reason behind resignation

Mrinank wrote in his letter that the world is in danger. This danger is not just from AI or biological weapons, but from many crises. He says that our capacity is increasing, but there is no balance with intelligence and understanding, otherwise there will be serious consequences.

He saw a disconnect between values ​​and work in the company. There is a lot of talk about AI safety on the outside, but under pressure on the inside, important things are ignored.

New beginning towards poetry

Mrinank now wants to leave technical work and study poetry. He considers poetic truth as important as scientific truth. He believes that the current crisis can be better understood and addressed through poetry and courageous speech. He wrote a poem by William Stafford "The Way It Is" Ended it with, which talks about sticking to the moral path.

Growing concerns in AI industry

This is not the first case. In 2020, Google’s Timnit Gebru left the company following controversy over her work on bias in AI. He later created his own organization on AI ethics and criticized the hypocrisy of companies. Mrinank’s resignation also highlights the growing gap between safety, values ​​and behavior in AI labs.


READ NEXT
Cancel OK