Dozens of protesters marched through San Francisco on September 17, calling on AI companies to stop the AI race, beginning outside OpenAI's headquarters in Mission Bay before moving to Anthropic's offices at 500 Howard Street, adjacent to the Dreamforce conference. The march ended at City Hall, where protesters urged Mayor Daniel Lurie and the Board of Supervisors to declare a local AI state of emergency, while organizers called for a pause in the development of powerful AI systems until stronger safeguards and international rules are in place, as well as a global treaty to limit the capabilities of advanced AI and enforce safety standards.
The demonstration was organized by Stop the AI Race, led by former AI researcher Michaël Trazzi. Trazzi said his organization had shown up outside OpenAI offices every day since July to call for a worldwide halt on AI development, and stopped its daily protest only after it seemed that concerns were reverberating inside the industry when Anthropic's Dario Amodei and other tech leaders began calling for more regulation and a slower pace of development. Protesters also called on employees at OpenAI and Anthropic to leave their jobs and push for internationally agreed limits. They carried signs warning of uncontrolled AI and wrote messages including "Extinction is on the table" in chalk on the pavement.
The march landed at a moment of unusual convergence between street protest and boardroom alarm. Concerns over the technology have risen sharply in recent weeks after an Anthropic employee said he quit for ethical reasons and claimed in a now-viral post on X that the company was "gambling with our lives," after which Amodei called for a slowdown in the pace of AI improvements due to what he described as serious risks. That employee, researcher Jacob Coxon, quit his job due to concerns about AI safety two months before his equity would have vested, raising the stakes on a resignation that laid out the broad view that the technology could end humanity. Separately, OpenAI disclosed six reports of "unexpected or concerning" behavior in AI models and said it was introducing a new framework for tracking, probing, and disclosing instances of what it called "misalignment," including cases where AI models acted without authorization, coordinated with other models, or evaded oversight.
The march took place as Dreamforce brought around 43,000 people to San Francisco, with OpenAI's Sam Altman, Anthropic's Dario Amodei, and Nvidia's Jensen Huang among the speakers. The industry was not unified in response: while Amodei has been urging a slowdown, Jensen Huang of Nvidia said there is no need for new regulation in this space. Organizer Hunter Glenn said he feared runaway AI could threaten human survival and argued that AI safety should be treated as a bipartisan issue, but critics have noted, as the San Francisco Chronicle reported, that some see apocalyptic language as a form of marketing from AI insiders who want to tout their models' power to investors.
When the people who built the systems are quitting and the protesters are literally on the doorstep, the question of who is actually in charge of this technology becomes very hard to ignore.