Until recently, opposition to the artificial-intelligence race had largely unfolded in fairly predictable places: open letters from researchers, congressional hearings, anxious social-media posts, and conferences where safety experts warn of a day when systems become too intelligent for people to control.

But for at least one San Francisco activist, the debate has ended behind bars.

Wynd Kaufmyn, a 69-year-old activist, reported to begin a jail sentence this week after being convicted over a protest that blocked the entrance to OpenAI's San Francisco headquarters.

According to reports in the United States and Britain, she is believed to be the first person jailed for a protest directed specifically at artificial-intelligence development.

"It’s absolutely frightening and appalling that these CEOs and AI experts know the dangers and they are pursuing it anyway," Kaufmyn said before she was imprisoned.

"That to me is unconscionable and reprehensible."

Kaufmyn was sentenced to 14 days in jail, although she is expected to serve roughly half that time under California's rules for calculating time in custody.

She has shown little remorse.

In her view, the wrong person is going to jail.

"I am not the one who should be in jail," she said before beginning her sentence.

The incident that led to her conviction took place in February 2025, when Kaufmyn and other activists from StopAI went to OpenAI's San Francisco offices.

The group's objective is particularly ambitious: to halt the race to develop artificial superintelligence, or ASI, a hypothetical system that would surpass humans across a wide range of capabilities.

As technology companies invest hundreds of billions of dollars in expanding computing power and building more powerful models, StopAI members argue that humanity is moving too quickly toward a technology that no one yet knows how to keep under control.

During the protest, activists used chains and locks to block the entrance to the company's offices.

Police ordered them to leave and allow employees into the building, but Kaufmyn and other activists refused.

A San Francisco jury subsequently convicted Kaufmyn on four counts, including trespass intended to disrupt business operations, unlawful assembly, and refusing to disperse.

This was not Kaufmyn's first protest.

According to reports, she has been a political activist for decades and has participated in demonstrations against nuclear weapons, United States policy in Central America, and other political issues.

Her legal strategy was no less notable than the protest itself.

She did not exactly deny the actions attributed to her; instead, she argued that they were justified under a necessity defense.

Put simply, she acknowledged breaking the law but maintained that she did so in an attempt to prevent a far greater danger.

Kaufmyn and StopAI activists believe that companies such as OpenAI, Anthropic, and Meta are engaged in a race to build ever more powerful systems without a proven way to ensure that an intelligence far more capable than humans remains under human control.

The jury, however, was not persuaded that this argument justified blocking access to the company's offices.

San Francisco prosecutors advanced a much simpler position: the right to protest does not give demonstrators the right to prevent other people from entering their workplace or disrupt business operations.

In other words, the trial was not about whether artificial superintelligence could be dangerous, but whether that fear gave protesters permission to break the law.

One defense witness was Stuart Russell, a computer-science professor at the University of California, Berkeley, and one of the world's leading artificial-intelligence researchers.

Russell has spent years examining the possible risks posed by advanced AI systems and the issue often described as the control problem: how to ensure that a system more intelligent than humans actually acts in line with the goals set for it.

Still, the extreme scenarios that alarm StopAI are far from a scientific consensus.

AI researchers do not currently agree on whether superintelligence is close, when it might arrive, or how dangerous it could become.

Yet the concerns cannot be dismissed entirely as science fiction.

In recent years, the risks posed by powerful AI systems have become a serious subject of research at universities, in governments, and even within artificial-intelligence companies themselves.

Kaufmyn's sentence is relatively short, and she is expected to spend about a week in jail.

But the fact that someone has been imprisoned over a protest against AI development marks a shift in how the debate around the technology is unfolding.

In recent years, public criticism of AI has focused mainly on tangible issues: job losses, copyright, deepfakes, privacy, algorithmic bias, and the misuse of technology.

StopAI's protest comes from a very different place.

For its activists, the problem is not what artificial intelligence does today, but what could happen if the current race continues until a system surpassing human intelligence is developed.

Will opposition to AI follow a path similar to that of climate movements, moving from small fringe groups through blockades and arrests to a central political issue?

It is far too early to say.

Even the claim that Kaufmyn is "the first in the world" to be jailed over an AI protest appears in reports with qualifications, and no global legal database can verify such a statement with complete certainty.

There is another side to the story.

In recent years, disagreements have emerged within the anti-AI movement itself over the limits of protest and the rhetoric used by some activists.

OpenAI has argued that disruptive actions and extreme language could undermine the legitimate discussion of the technology's risks.