‘Scream if you want to move slower!’ A nascent AI protest coalition comes together in London

A diversity of motivations present both a challenge, and an opportunity, for those protesting AI

Protesters March near London’s King’s Cross. Credit: Pull the Plug

“Look around you,” PauseAI Global’s CEO Maxime Fournes told protesters outside Google DeepMind’s headquarters on a chilly day in London late last month. “Look at who’s here today. We do not agree on everything. We come from different organizations, different backgrounds. We have different concerns: safety, ethics, labor, democracy, art, human dignity.

“We’ve all arrived at the same conclusion: slow down. When this many different people from this many different perspectives all say the same thing, maybe it’s time to listen.”

The protest march past the London offices of OpenAI, Meta, DeepMind and Google culminated in an “assembly” in which participants were encouraged to discuss their fears about AI in small groups. Respectful disagreement was welcomed.

In the diversity of attendees and subjects discussed throughout the day, the protest showed some early indications that bridges are being built which could lay the foundations for a mass anti-AI movement. As well as being the biggest AI protest to take place in the UK, the event also marked the entry onto the scene of a new group: Pull the Plug.

That campaign was established in the UK at the end of last year by Harry Atkinson, a former story editor in the film industry, and Frieda Lurken, a co-founder of climate group Extinction Rebellion. Explicitly set up to be a popular movement, the group calls for legally binding Citizens’ Assemblies deciding how AI should be developed. Such solutions are a common feature of previous movements focused on issues such as climate and economics, though they have rarely made much headway. Given the lack of even the most basic AI regulation in the US and UK to date, assemblies may face an uphill battle to exert much influence.

That, however, may not get in the way of building public pressure by bringing together different movements. Pull the Plug worked to organize the London event with PauseAI, the international AI safety group established in 2023 to call for a pause in frontier AI development, which already has an active chapter in the British capital.

The event was attended by an estimated 300 people at its peak, with Pull the Plug putting the total for the entire day, including the assembly, at 500. Those numbers are tiny compared to some of the massive protests for other causes London has witnessed in recent years. But it was far larger than a previous PauseAI protest last summer, which attracted just a few dozen people.

“What seemed like a tiny protest eight months ago was leading to this, and this will be tiny compared to what comes next,” PauseAI UK director Joseph Miller told me.

---

Will AI safety become a mass movement?

---

Movements against AI often fall into one of two camps: safety campaigners concerned about existential threats to humanity, and ethics campaigners worried about immediate harms. Lurken told me that “there’s good points on both sides”.

“Ultimately it doesn’t have to be a decision between those two sides and we can talk about the ways in which AI is impacting our lives right now already, and we can talk about some of the less likely but very serious scenarios of what could happen in the future. And the two things don’t have to be played out against each other.”

To make an impact, she said, the different sides needed to come together.

Several of the people I spoke to described AI usage as something mostly put to frivolous uses, often giving the creation of memes as an example. Combined with fears about how using the tools could be detrimental to cognitive development, these protesters were skeptical that most people need or would benefit from AI tools — or that they would ever be worth the tradeoffs. One musician I spoke to, Andy, suggested a kind of licensing scheme: instead of an all-out ban, only legitimate uses of AI would be authorized.

On the ground, one of the most salient points was the environment. Attendees frequently cited data centers and water usage as a concern. This overlap is hardly surprising: organizers cited support for the protest from Global Action Plan, a group which coordinated several demonstrations against data centers across the UK over the same weekend.

---

We’re getting the argument about AI's environmental impact all wrong

---

The week of the protest, headlines had been dominated by Anthropic’s battle with the Pentagon and OpenAI’s subsequent replacement as the US military’s classified systems AI provider. I had expected that topic to come up in every conversation at the protest. Yet those I asked about it hadn’t been following the drama. The call to attend had reached well beyond the kind of person who keeps tabs on every AI policy and business update.

The dispute was mentioned in one early speech given by Irina Tavera, organizing director of PauseAI, who emphasized the dangers of autonomous weapons. The issue could yet be a catalyst for building a broader movement: new research from Social Change Lab seen by Transformer found that autonomous weapons are the issue most likely to make the public view AI as a threat.

---

OpenAI’s Pentagon red lines are a mirage

---

Setting off from OpenAI’s discreet London offices, the march processed through the area surrounding King’s Cross Station, briefly stopping the traffic on a major artery of the capital. A cyclist blasting Daft Punk brought up the rear. Diners in a nearby upmarket burger restaurant paused their lunch for a moment to gape as people streamed by.

At each stop Lydia, the rabble-rousing emcee, would warm up the crowd before handing over to guest speakers. Those chosen to address the protesters at various points along the route reflected the mix of concerns. While PauseAI’s leadership team spoke of existential risk, others talked of different impacts AI could have.

Tapping into a more general feeling of techlash was Dr Rachael Kent, a lecturer in Digital Economy and Society Education at King’s College London. She won a landmark class action against Apple in the UK last year over its App Store pricing tactics, which she said “demonstrated that even tech monopolies are not above the law, and they can — and they must — be held to account.” Her speech hardly touched on AI, instead zeroing in on the issue of corporate power.

Outside the offices of Meta, Lydia led the crowd in a call-and-response chant about whether they trusted the tech giants in this new technological boom, considering the failings of the social media era. “Do we trust them to protect our jobs? Do we trust them to protect our data? Do we trust them to protect our children?” A resounding “no!” was the answer each time.

The impact that AI could have on the creative industries — or even creativity itself — was also front of mind for many of the protesters I spoke to.

“Amongst our community, a lot of people are very, very concerned because AI is just scraping our work and regurgitating it and replacing us,” one woman, who had come as part of a trio of picturebook authors and illustrators, told me. She had spent six hours making a flag for the occasion, bearing the phrase “AI is the thief of everything.”

“Maybe there will come a time when we won’t be able — we won’t have to make work,” her friend added. They were worried about the “crushing of creativity” and how this would affect humanity more widely.

PauseAI’s Tavera used the moment to draw a parallel between this moment and the dawn of nuclear weapons, arguing that global cooperation was needed. “We know that states, nations, can choose coordination over competition.”

Even as the event welcomed newcomers to the cause of AI safeguards, there were still participants who hailed from areas more traditionally associated with existential risk safety campaigners: academia, technology, policy.

Early on, I spoke to Callum Burns-Curd, who has a Masters in Computational Intelligence and was skeptical that models yet have anything like catastrophic capabilities, but had still been alarmed by how rapidly the technology has developed. When I asked him what his main concern was, he cited being turned into Nick Bostrom’s proverbial paperclip. He added, in understated British fashion: “I’m not a luddite, but I wouldn’t like to promote the extinction of humanity.”

The diversity of motivations on show among the attendees present both a challenge, and an opportunity, for the organizers. Whether it’s the end of artistic creativity, or mass extinction, a broad church will be required to create the kind of mass movement that can influence, let alone stop, AI. Holding that coalition together, and expanding it, will be their next test.