1月6日组织者艾米·克雷默领导保守派反对AI数据中心

Oliver Kemp for Transformer

“I didn’t know anything about Amy. I didn’t know she helped organize January 6 — I just learned that today from a media person. My focus is on opposing the data center.”

Renee Spaziani was shouting over the sounds of vuvuzelas and chants at a protest outside a municipal building in central New Jersey, which she had helped organize in response to an ad on Facebook. “I was like, ‘I don’t know if I want to do this.’ And I saw it a second time. I was like, ‘we need more eyes on Kenilworth. We have to do this.’”

Spaziani was one of a group of Union County residents pushing back against the construction of local data centers at borough council meetings and small demonstrations. Amid a hellish mix of Canadian wildfire smoke and flash flood-triggering showers last month, they were joining forces with another group called Humans First to stage a “National Day of Protest”.

Cut through the noise.

Spaziani had earlier received an email from some prospective protesters concerned that Humans First was heavily aligned with conservative movements. They were progressives, they explained, and worried the organization also stood for things they vehemently opposed, such as white nationalism, ICE, war, and attacks on women’s rights. Spaziani brushed it off — her political allegiances lie with neither the right nor the left, she said — but the issue was coming up again as journalists asked if she knew anything about the national organization co-hosting her event.

Spaziani did her best to refocus attention on the cause at hand. The company they were protesting, CoreWeave, had “done a significantly amazing job connecting people from all different walks of life” in opposition to their data center, she said.

The Amy in question was Amy Kremer, chair of Humans First and a longtime conservative activist and Trump supporter — who, as Spaziani learned that day, did indeed help organize January 6. Now, though, she has a new focus: mobilizing a conservative coalition against AI companies.

“I know how to organize people,” Kremer told Transformer. “I’ve done it time and time again, and that’s why I’m a threat to their power.”

Kremer pictured with Trump in the Oval Office. Credit: Humans First

Amy Kremer helped procure the permit for the rally which preceded the January 6 riots, though she did not participate in the attacks on the Capitol. Self-described as the “founder of the modern day Tea Party,” she was a leading denier of the 2020 election results, and has been a prominent figure in right-coded conspiracy theory movements around issues such as COVID and President Obama’s citizenship. She was recruited this year to to Humans First, then a nonpartisan organization aiming to mobilize a grassroots movement on AI safety, to help reach conservatives. In April she became its chair.

As a group, conservatives are only slightly less concerned about AI than Democrats. A 2025 Pew Research poll found that 51% of Democrats and 50% of Republicans were more concerned than excited about the increased use of AI in daily life. Another recent poll from the AI Policy Institute cited by Kremer showed that 77% of Republicans said they wanted to see regulatory guardrails on the technology, compared to 82% of Democrats. The conclusion for many in the AI safety movement has been that conservatives are primed to get behind the issue, and just need the right leader to organize them.

Yet a class of investors and entrepreneurs who want to see as few guardrails as possible have pushed Republican policy in the opposite direction. Many of these “accelerationists” hold or have held roles in the Trump administration, including venture capitalists Marc Andreessen, Sriram Krishnan and David Sacks. There are signs their influence is waning — Sacks and Krishnan have stepped back in recent months, Sacks to an “advisory” position, Krishnan altogether — and the administration has shifted towards a more hands-on approach, such as implementing export controls on what it sees as dangerous models. But still, their proximity to the president has yielded a generally pro-industry platform that prioritizes innovation over caution, making it difficult for more wary conservatives to take a stand against AI.

Kremer’s resume makes her loyalty to the president difficult to question, and she says she believes she’s figured out an approach that works: “I know the president can be moved on issues.” Big AI firms have bought a lot of influence, she says, but politicians are meant to represent people and grassroots activism can tip the scales.

“Big Tech has all the money. They bought off everybody in Washington DC — I mean, they built a moat around the White House. They bought off people at the state level. Now it’s down to the county commission level. They have bought off everybody they possibly can,” says Kremer. “We will never be able to compete with their money. But what they cannot buy off is the people.”

Conversations about what it would take to organize conservatives around AI safety have been happening for years, says Joe Allen. A concert rigger-turned-conservative ideologue, Allen was working as a technology writer, then as an editor on Steve Bannon’s War Room, when AI became his primary focus. He is particularly concerned with “transhumanism,” a belief that technological innovation can transcend the travails of the human condition. Transhumanism is all too common among the Silicon Valley leaders driving the AI boom, Allen says, and a threat to society. That should particularly concern Christians, he says, because it asserts that human invention can supersede God’s creation. He summarized his beliefs about AI as early as May 2025 as “Humans First.”

It was around that same time, according to Allen, that he began thinking about ways to bring people like him — the MAGA crowd who tuned into his segments on War Room, but also disaffected liberals and others eager to reject establishment narratives — into the fold. Researchers and AI advocacy groups were talking about how to regulate AI, but those conversations were primarily driven by AI pessimists Allen collectively refers to as “doomers.” He was never convinced AI would lead to collective ruin as they do, he says, but found some overlap between their concerns and his own. He also found their warnings about AI compelling enough to be the basis for a social movement, even if exaggerated. Conservatives could be turned, if only someone explained it to them, Allen and the doomers agreed.

“It’s not getting people to wake up about AI at this point — that’s long since passed. It’s getting them to wake up to the false narratives, exaggerated narratives, manipulative narratives,” Allen says.

Allen pictured on an episode of War Room with Steve Bannon. Credit: warroom.org

In 2025 Allen met Oliver Zhang and Dan Hendrycks, managing director and executive director at the Center for AI Safety (CAIS), and Allen says talks about creating Humans First with them started last summer. The non-profit does technical and policy-minded AI safety research, and also has a 501(c)(4) lobbying arm and a PAC (which is now inactive). In January this year, Zhang and CAIS special project associate Arunim Agarwal incorporated in California what would become Humans First. According to NBC, Humans First was “incubated” by CAIS, which also gave the group an “initial loan that would be paid back after the group raised donations from individuals who cared about fostering safe and beneficial AI.”

At that time, Humans First was pitched as nonpartisan, with left and right “coalitions.” Kremer led the right wing; the left was run by progressive organizer Alexander McCoy. It was launched publicly in March.