On Saturday, roughly 200 demonstrators marched through San Francisco, moving from Anthropic’s offices on Howard Street to OpenAI’s headquarters on 3rd Street and then to xAI on 18th Street. The march was organized by Stop the AI Race. Its central demand: that the CEOs of all three companies publicly commit to pausing development of frontier AI models — if their competitors agree to do the same.
The march was organized by Michaël Trazzi, a former AI safety researcher turned documentary filmmaker. In September 2025, Trazzi staged a three-week hunger strike outside Google DeepMind’s London office with a similar demand. The protest drew results of a sort: at a Davos forum in January 2026, DeepMind CEO Demis Hassabis said he would be open to a conditional pause, provided all parties in the race agreed. Anthropic CEO Dario Amodei said around the same time that he would meet with Hassabis “right now” — if only the two of them were involved.
Saturday’s crowd included researchers, academics, and activists from the Machine Intelligence Research Institute, PauseAI, QuitGPT, and other organizations. Speakers included MIRI president Nate Soares, Université de Montréal professor David Krueger, and UC Berkeley statistics professor Will Fithian.
The protest had a specific backdrop. In February 2026, Anthropic quietly walked back a core commitment in its Responsible Scaling Policy — a promise not to train models more powerful than its current ones without verified safety guarantees. The updated policy replaced that hard rule with advisory guidelines. Two weeks earlier, OpenAI had signed a contract with the Pentagon, a move that galvanized many in the crowd.
The march ended at Dolores Park. No incidents were reported.