San Francisco AI Firms Lose Two Top Researchers Amid Safety Concerns

Two of San Francisco County’s highest-profile AI researchers resigned on February 12, 2026, publicly airing concerns that strike at the heart of local AI labs’ safety claims. Mrinank Sharma, a former safety researcher at Anthropic, shared a “lengthy letter” saying “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,” while Zoë Hitzig, who had worked at OpenAI for two years, published a New York Times essay announcing her resignation and criticizing the company’s approach to privacy and safety.

Hitzig’s New York Times essay, reported by Finance Yahoo and CNN, said she had “deep reservations” about OpenAI’s emerging advertising strategy and warned that ChatGPT’s archive of user data included users’ “medical fears, their relationship problems, their beliefs about God and the afterlife.” Nationaltoday reproduced Hitzig’s line: “I once believed I could help the people building A.I. get ahead of the problems it would create. This week confirmed my slow realization that OpenAI seems to have stopped asking the questions I'd joined to help answer.”

Sharma’s departure came from Anthropic, which Nationaltoday identifies as a San Francisco-based company positioned as focused on AI safety and ethics. Nationaltoday summarized Sharma as publicly voicing concerns about Anthropic’s ethical practices and reported he described struggling to uphold the company’s values in his letter; the outlet’s San Francisco top stories page carries a Feb. 17, 2026, date stamp.

Industry reporters say the departures are part of a broader wave of resignations across the AI sector. Finance Yahoo and CNN reported churn at xAI, where two co-founders quit within 24 hours and at least five other xAI staff announced departures on social media over the past week. Finance Yahoo cites Elon Musk’s post that xAI was “reorganized” to speed up growth, which “unfortunately required parting ways with some people.”

Broader commentary and online analysis add context but vary in sourcing. A Medium post by Raybwilliams cites an MIT Media Lab study it says links heavy reliance on AI to “cognitive atrophy,” quotes Woxsen University on an “evolutionary mismatch,” and reports a finding that “67% of young adults in major cities can’t give directions to familiar locations without consulting their phones.” The Medium piece also cites the Future of Life Institute AI safety index as saying leading firms including OpenAI, Anthropic, xAI, and Meta “are severely underperforming relative to emerging global standards.”

Community reaction in Silicon Valley forums has been sharp. A Hacker News thread reproduces a user “tim-tday” who wrote: “The problem is the system. Enshitification will happen because we push our institutions to make a profit at all costs… AI companies exist in the former system and they will all sacrifice their morals and betray us.” Nationaltoday noted that these departures will likely prompt increased scrutiny from the public and regulators.

Key follow-ups remain: obtain Sharma’s full letter and its date, publish Hitzig’s complete New York Times essay with its publication date, and seek formal responses from OpenAI, Anthropic, and xAI about the criticisms. Reporters should also verify the MIT Media Lab study and the Future of Life Institute index cited on Medium, and confirm identities and dates for the xAI co-founder departures. These resignations, the public quotes from Sharma and Hitzig, and the reported xAI churn leave San Francisco’s AI scene facing intensified local and regulatory scrutiny.