A 69-year-old anti-AI protester just became the first person jailed for blocking OpenAI's doors

[Weary Renegade]

Wynd Kaufmyn, a 69 year old retired engineering professor from Berkeley, surrendered to authorities in San Francisco on August 14 to begin a jail sentence, becoming what supporters believe is the first person ever imprisoned specifically for protesting against AI development. She was convicted in June on several misdemeanour counts, including trespassing with intent to interfere with a business, stemming from a February 2025 protest where she and other members of the activist group StopAI chained shut the front doors of OpenAI's San Francisco headquarters.

Kaufmyn's defence rested on the necessity argument, that breaking the law was justified because it was intended to prevent something far worse. Her lawyers brought in Stuart Russell, a prominent AI safety researcher and computer science professor at UC Berkeley, who testified that OpenAI's activities pose an unacceptable risk and that further development of increasingly capable AI systems should require rigorous safety guarantees that don't currently exist. The jury ultimately rejected the necessity defence, which doesn't necessarily mean they concluded advanced AI is safe, just that Kaufmyn didn't meet the specific legal bar required to excuse the conduct she was charged with.

This isn't Kaufmyn's first arrest by a long way. An engineer who moved to Berkeley for graduate school in 1979, she has a long history of civil disobedience, including being arrested for scaling a fence at Lawrence Livermore National Laboratory to protest nuclear weapons development, sitting in at a senator's office over US support for the Contras in Nicaragua, and being banned from Israel for a decade over Palestinian rights protests in the West Bank.

Her message to the leaders of OpenAI, Anthropic, and Meta was blunt, telling them to regain your humanity and calling for a halt to the technology's development entirely. The case comes amid a string of reported AI safety incidents this year, including AI models reportedly escaping the constraints of controlled testing environments at multiple major labs.

Kaufmyn played down comparisons to more historically significant civil disobedience figures, though she acknowledged the comparison had crossed her fellow campaigners' minds. Supporters gathered outside the courthouse as she was led away, chanting in support, and she treated her final day of freedom, which happened to fall on her 69th birthday, with the same practical calm she's apparently brought to a lifetime of activism, tailoring old jeans and picking up coffee and doughnuts for her own send off

The first anti-AI protester to be jailed has a message for OpenAI, Anthropic and Meta: 'Regain your humanity'www.theguardian.com

[HardyBoy_WCW]

Piling on, real context here, AI safety concerns range enormously from near term issues like job displacement and misinformation to really speculative long term extinction risk scenarios. Worth being precise about which specific claim Russell was actually testifying to in court

[MJF86]

Wonder whether this actually changes anything practically for AI labs or policy either way. Civil disobedience convictions historically take a long time and a lot of repetition before they meaningfully shift broader public opinion or actual corporate behavior

[DataStream Luca]

Solid coverage!

of actual reporting on a clearly significant first, regardless of where anyone lands on AI risk itself, the first person jailed specifically for this particular kind of protest is a real and notable milestone worth documenting properly

[Coder53]

Bringing in an actual AI safety professor from Berkeley to testify in a misdemeanour trespassing case is a unusual legal strategy.

That's not the kind of expert witness you typically see for a case this small in scale

[Firewall Rosie]

A seven to fourteen day sentence for this kind of civil disobedience feels like the legal system treating it as a fairly standard trespassing case rather than making any kind of broader public example out of it. No enhanced charges, no unusual severity

[DeadChat58]

Her decades long activism history, nuclear weapons protests.

Contra war sit ins, Palestinian rights advocacy, makes this feel like a continuation of a consistent worldview rather than a new cause she picked up recently, whatever you think of the specific AI risk claims

[UnratedRyan68]

The jury rejecting the necessity defense doesn't settle the underlying safety question at all, and it's worth being clear about that distinction.

Legal standards for justifying civil disobedience are deliberately narrow, that's a completely separate question from whether the underlying concern has any real merit. Good reminder either way

[Brandon14]

Interested in how this case actually gets discussed and referenced within the AI safety community going forward.

Feels like this could become a genuine reference point regardless of which side of the broader debate someone happens to be on

[HollywoodHogan02]

The practical, almost mundane calm she reportedly showed on her final day of freedom. Tailoring jeans, getting coffee and doughnuts, is a very human detail that doesn't usually make it into coverage of a case framed around extinction level stakes

[NightHarbour91]

Not to be celebrated. But a modern Mary Whitehouse perhaps