Local
Four Arrested as Anti-AI Protesters Block Anthropic Headquarters on IPO Day
More than 100 demonstrators gathered outside Anthropic's San Francisco offices Thursday as the company went public in the largest IPO on record, with four arrested after blocking the entrance. Protesters demanded a ban on recursive self-improvement and warned of existential risk from frontier AI.
SAN FRANCISCO â Four people were arrested Thursday outside Anthropic's headquarters after demonstrators blocked the entrance to the building at 500 Howard St., timed to the artificial intelligence company's record-setting debut on the public markets.
More than 100 protesters rallied outside the office, voicing fears that increasingly powerful AI systems pose a threat to humanity. Their concerns ranged from the possibility of an AI market bubble to worries that labs are deploying models they cannot fully control to accelerate their research.
Around noon, dozens of demonstrators sat down or stood in the doorway of the company's offices. When private security could not persuade them to leave, San Francisco police arrived and made arrests at about 2:30 p.m. The department confirmed that arrests took place but said it could not confirm the number of people taken into custody or the charges because the investigation is ongoing.
Among the speakers was Jacob Coxon, a former Anthropic researcher who publicly resigned from the company in September. Coxon has criticized AI leaders as irresponsible, accusing them of "gambling with our lives."
Organizers focused much of their anger on recursive self-improvement â the idea that an AI system could autonomously design and build its own successor. Anthropic, the maker of the Claude chatbot, states on its website that such technology is not inevitable but could arrive "sooner than most institutions are prepared for."
"Anthropic is the corporation that is the closest to achieving recursive self-improvement," said Phoebe Thomas Sorgen, one of the protesters blocking the entrance. "It is already out of control, so we have to stop it. Now, we have to take nonviolent direct action at the location of the crime. This is a crime scene."
David Krueger, an AI professor at the University of Montreal, also addressed the crowd.
The action coincided with Anthropic's initial public offering, the largest in history. Company shares opened at $111.50 Thursday before closing at $97.17. The offering has stirred a mix of excitement and anxiety across the Bay Area, where residents already contending with soaring rents and living costs are weighing how a newly public AI giant will reshape the region.
Similar demonstrations unfolded earlier in the day outside the Nasdaq exchange in New York, where protesters were arrested ahead of the listing. It was at least the second time in a week that demonstrators had gathered outside Anthropic's Bay Area office.
Protesters said they want lawmakers to impose industry-wide rules to protect the public from the consequences of further AI development.
"It seems pretty clear to me that the next step is for governments to step in and say, 'Hey, you can't develop technology that you say will greatly endanger everyone's lives and national security. So cut that out,'" said Jeffrey Ladish, executive director of Palisade Research, a nonprofit that studies AI safety.
Thomas Sorgen put the stakes in blunter terms, arguing that the industry's own leaders acknowledge the danger.
"Those CEOs are admitting that their race toward superhuman artificial intelligence threatens human extinction â significantly," she said.