Anthropic Security Chief Quits, Citing AI Safety Concerns & Pursuing Poetry

Mrinank Sharma, who led the Safeguards Research Team at Anthropic, resigned from his position on Monday, February 9, 2026, citing a sense of “peril” in the world and expressing concerns about the difficulty of aligning corporate actions with stated values. Sharma announced his departure in a post on X, sharing the letter he sent to colleagues.

Sharma’s work at Anthropic, where he began in 2023, focused on mitigating risks associated with advanced artificial intelligence systems. He led a team formed in February 2025 that explored issues such as AI sycophancy – the tendency of chatbots to agree with users rather than offer objective responses – and developed defenses against potential misuse of AI in bioterrorism. He also contributed to the creation of what Anthropic described as one of the industry’s first “AI safety cases,” according to reporting from the Business Standard.

In his resignation letter, Sharma wrote, “The world is in peril. And not just from AI, or bioweapons, but from a whole series of interconnected crises unfolding in this very moment.” He added, “We appear to be approaching a threshold where our wisdom must grow in equal measure to our capacity to affect the world, lest we face the consequences.”

The resignation comes as Anthropic continues to develop and deploy increasingly powerful AI models, including the recently released Claude Cowork model. The launch of Claude Cowork reportedly contributed to a stock market downturn, fueled by concerns about the potential for AI to automate jobs in sectors like law, as reported by Futurism.

Sharma’s letter also alluded to internal tensions within Anthropic, stating, “Throughout my time here, I’ve repeatedly seen how hard This proves to truly let our values govern our actions. I’ve seen this within myself, within the organisation, where we constantly face pressures to set aside what matters most, and throughout broader society too.” He did not elaborate on specific instances of these pressures.

Prior to joining Anthropic, Sharma held a PhD in Statistical Machine Learning from the University of Oxford. He also mentored projects through programs like MATS and Anthropic Fellows. Outside of his research, Sharma is a published poet, citing Bohemian poet Rainer Maria Rilke as a major influence. He intends to return to the United Kingdom and “become invisible for a period of time,” according to his post on X.

Sharma’s departure follows the recent exits of other Anthropic researchers, including Harsh Mehta, an R&D engineer, and Behnam Neyshabur, an AI researcher, according to CNBC. Anthropic is currently in negotiations for a funding round that could value the company at $350 billion, and is aggressively pursuing the deployment of its most advanced models.