Stop the AI Race

Stop the AI Race is a US-based grassroots movement founded by Michaël Trazzi — a French filmmaker and former AI safety researcher — to pressure the leadership of major AI laboratories into committing to a coordinated pause on frontier model development. The organization's single formal demand: every major AI lab CEO must publicly commit to pausing frontier model development if every other major lab in the world credibly does the same. The conditional structure is designed to short-circuit the competitive-pressure rationale that individual labs cannot slow without coordination — by seeking a verifiable mutual pledge rather than unilateral restraint. Stop the AI Race has organized two major protest marches in San Francisco and a sustained direct action at OpenAI's headquarters, establishing itself as a leading coordinator of street-level AI moratorium activism in the United States.

Founding and the London hunger strike

The movement cohered around a hunger strike Trazzi conducted outside Google DeepMind's London office beginning 5 September 2025, lasting approximately one week before he ended the action on medical advice after fainting. The demand addressed directly to Google DeepMind CEO Demis Hassabis: publicly commit that DeepMind would halt the development of frontier AI models if all other major AI companies in the West and China did the same. Denys Sheremet of Together Against AI joined Trazzi at the London office and continued the strike after Trazzi withdrew. The London action ran concurrently with a parallel hunger strike by Guido Reichstadter outside Anthropic's San Francisco headquarters. Trazzi has cited philosopher Nick Bostrom's Superintelligence as foundational to his thinking, and frames the competitive-pressure coordination problem — AI labs claiming they cannot pause unilaterally — as the specific target of the conditional demand.

March 2026 San Francisco march

On 21 March 2026, Stop the AI Race organized its first major US protest action: a march in San Francisco from Anthropic's offices to OpenAI's to xAI's, drawing approximately 200 participants. The coalition included the Machine Intelligence Research Institute, PauseAI, Stop AI, QuitGPT, and Evitable. Speakers included Holly Elmore (then Executive Director of PauseAI US), Dr. David Krueger (AI professor at the University of Montreal and founder of Evitable nonprofit), Nate Soares (MIRI President), and Will Fithian (UC Berkeley professor). The march publicly named Dario Amodei (Anthropic), Sam Altman (OpenAI), and Elon Musk (xAI) as the target recipients of the conditional-commitment request. The action was reported in the New York Times, the Washington Post, and The Atlantic.

July 2026 march — largest AI protest in US history

On 11 July 2026, Stop the AI Race organized the Stop the AI Race march (San Francisco, 11 July 2026), drawing approximately 400 participants — described as the largest AI protest in American history to that date. The thirteen-organization coalition included PauseAI, Stop AI, QuitGPT, Evitable, AI Action, the National Union of Healthcare Workers, and Food and Water Watch. The march ran noon to 3pm from OpenAI's Mission Bay headquarters (1455 3rd St) to Anthropic's SoMa offices (500 Howard St) to Google DeepMind's Rincon Park location (345 Spear St), with rallies and speeches at each stop. Organizers cited Anthropic's 2026 decision to drop its Responsible Scaling Policy — which had committed the company to pause development if its AI became too dangerous — as a proximate cause for the march's timing and centering of Anthropic as a named target. David Krueger served as co-organizer and speaker alongside Trazzi.

Occupation and ongoing direct action

On 22 July 2026, Stop the AI Race began a continuous occupation outside OpenAI's San Francisco headquarters — a form of sustained presence designed to maintain public pressure on company leadership. The occupation later shifted to New York City. The movement has operated as a central coordinator in US AI moratorium organizing, regularly convening coalitions with PauseAI and Stop AI while maintaining a distinct organizational identity anchored to the conditional-pause demand and Trazzi's direct-action orientation.