StopAI Cofounder’s Disappearance—and the Warning Signs of Radicalization

On November 22, StopAI—an Oakland-based, “nonviolent civil resistance” group seeking a permanent ban on the development of artificial intelligence—posted on X that its cofounder, Sam Kirchner, was missing. Earlier in the week, the group said, Kirchner had assaulted another member who refused to let him purchase a weapon with organization funds. He had also made violent statements that led others to fear he intended to use the weapon “against employees of companies pursuing artificial superintelligence.” Two days later, he vanished from his apartment. Shortly afterward, police responded to a 911 call about a man making threats near one of OpenAI’s San Francisco offices. It remains unclear whether this was Kirchner; as of this writing, his whereabouts remain unknown to the public.

Fears about the “existential risks” posed by artificial intelligence have circulated in Silicon Valley for years, and observers like Nirit Weiss-Blatt and Dean Ball have flagged StopAI as a particularly incendiary voice in these warnings. But here, in the fugitive Kirchner, we may have something new: an actual would-be assassin, perhaps armed, and reportedly motivated to inflict harm on the leaders of American frontier AI labs.

The latter point is important. Kirchner is not the first Bay Area AI obsessive to take up arms and issue credible threats of violence. Earlier this year, a cult of trans-identified individuals known as “the Zizians” went on a murder spree that began with an 80-year-old landlord in the East Bay and ended in a shootout with Border Patrol agents in Vermont. Ziz, the group’s eponymous leader, supposedly viewed the risks posed by AI as “the most important problem in the world” and was fixated on Roko’s Basilisk—a thought experiment in which an omniscient superintelligence punishes anyone who knew it might one day exist but failed to help bring it about. But the Zizian murders were not political assassinations; they were, it appears, expressions of petty grievances or general psychosis, unconnected in any immediate way to AI-related activism.

So Kirchner—who has publicly stated that OpenAI’s executives should be “charged with attempted murder of eight billion people” and is now allegedly trying to fight for humanity himself—marks an apparent escalation in anti-AI radicalism, albeit one years in the making. Despite its stated commitment to nonviolent resistance, StopAI’s protests and rhetoric are steeped in references to death and murder. Since April 2024, the organization has held regular “Stop AI or We’re All Going to Die” protests outside the offices of OpenAI and Microsoft. At one such protest, a person dressed like a robot emblazoned with the OpenAI logo pretended to shoot a handcuffed man in the head. According to its website, the group seeks a ban on the development of AI specifically “to prevent human extinction in 1-3 years.” And on a recent podcast, Kirchner said he and other StopAI members don’t care about their criminal records because “we’re all going to be dead soon.”

In Kirchner and StopAI, we find allied the near-exact conditions necessary to create a righteous assassin. Kirchner, a former DoorDash driver and electrical technician, appears to be a disaffected young man fixated on abstract visions of extreme violence. StopAI, and the broader fringes of the anti-AI movement, provide the ideological gloss that gives this fixation a moral rationale. (Their specific arguments for why AI will imminently kill us all are too involved to revisit here, but the basic idea is that a sufficiently “intelligent” system could recursively self-improve until it surpasses human capabilities and thus becomes able to destroy us.) The only missing ingredient, it seems, is a circle of supporters willing to look the other way as a friend turns violent. To StopAI’s credit, the organization publicized Kirchner’s disappearance and alerted police to his stated intentions.

We can be grateful that Kirchner’s threats remain unrealized. But we should stay alert to the warning signs of radicalization: a disaffected young person, consumed by abstract risks, convinced of his own righteousness, and embedded in a community that keeps ratcheting up the moral stakes. None of these conditions is unique to anti-AI activism. They echo, to take one example, the psychology of extreme climate change obsessives, some of whom swear off having children and idolize those who self-immolate in protest. Can we pull those who flirt with violence back from the abyss? We should hope so. Otherwise, many dark days lie ahead.

Photo by John Ricky/Anadolu via Getty Images)