One anti-data-center group's funding really does trace back to Silicon Valley money. The "bused-in protesters paid by Beijing" story fell apart the moment a reporter asked attendees how they got there.
Box Elder County commissioners approved a 9-gigawatt data center campus near Tremonton, Utah on May 4, 2026, over hundreds of protesters. Lead developer Kevin O'Leary said more than 90 percent of the protesters were paid and bused in from out of state, and floated China as the only beneficiary of the opposition. Commentator Jordan Schachtel and the youth group Bull Moose Project separately traced the advocacy coalition Humans First to funding linked to Open Philanthropy. By mid-July, White House adviser David Sacks and the All-In podcast were calling the broader national data center protest wave a "psyop."
The specific claim about Humans First's money holds up in part. Reporters at NBC and the pro-buildout Bull Moose Project both traced the group's origins to a Center for AI Safety project with Open Philanthropy-linked funding, though organizer Amy Kremer denies taking Moskovitz money now or ceding control. That's a real, documented thread worth pulling on.
The bigger claims don't survive contact with reporting. KSL asked rally attendees directly and found unpaid Utahns who drove themselves, not a paid bus fleet. National polling shows broad, organic opposition to data centers, not a manufactured minority. And nobody, including O'Leary, has produced actual evidence for the China link. Mixed evidence: one group's funding history is fair game, the mass-inauthenticity story isn't.
Three of the four predictions behind this claim fell apart. The fourth was a mixed bag.
Being right about what happened and wrong about what it proves are different failures, so these two lines never get averaged into one score. How this works
The logic check
If the claim is right, these things should be true. Here's what we found. Each check says which version of the claim it's testing: the plain one most people would hear, or the widest one the same words could carry.
If most Tremonton protesters were paid and bused in from out of state, later rallies should turn up nonresident or compensated attendees.
FalseAt a May 23 Capitol rally, KSL asked attendees directly. Everyone interviewed said they were an unpaid Utahn who paid their own gas money to show up.
If the funding-lineage claim is accurate, reporting should document early effective-altruist money and staff behind Humans First.
MixedNBC's reporting and the Bull Moose Project both trace the group's origins to a Center for AI Safety project with Open Philanthropy-linked funding. Organizer Amy Kremer denies taking Moskovitz money or ceding current control, so the origin story checks out while the present-day claim is disputed.
If the opposition were manufactured rather than organic, national polling shouldn't show broad independent hostility to data-center siting.
FalseA June Reuters/Ipsos poll found only about a third of Americans approve of the pace of data center construction, and just 14 percent would want one in their own community.
If Chinese influence were actually driving the protests, carriers or officials should be able to point to real attribution, like platform takedowns or an intelligence finding.
FalseFortune notes O'Leary's paid-and-bused claim was made without evidence, and no one in the carrier set produced any actual attribution for the China claim beyond the talking points sounding similar.
The receipts
Every quote, checked against its source.
David Sacks, White House AI adviser
"a censorship power play"
Kevin O'Leary, lead developer of the Stratos project
"Which adversary would want that? There's only one: it's China."
Jordan Schachtel, The Dossier
"Humans First, a political astroturf group founded in San Francisco"
Nathan Leamer, executive director of Build American AI
"talking points are now indistinguishable from those of Global Times"
The All-In Podcast (David Sacks and co-hosts)
"The Datacenter Psyop"
Bull Moose Project (report conclusion, via Washington Examiner)
"The pattern is consistent: hire enough Republicans to claim bipartisanship while the operational staff, donors, and policy agenda remain overwhelmingly Democratic"
The full scoring
The raw numbers behind the verdict. What R/I/P/E/L and charge stages mean: how we score.
Research verdict: mixed evidence. This is the category the underlying research assigned before we graded the individual checks, and it came back the same on every card in this set.
Reach 4 for carriage spanning a White House adviser's 21-million-view post, a million-download podcast, and major-outlet reporting of the claims, short of saturating major-outlet adoption as the outlets' own frame. Intensity 3 for bad-faith default and motive attribution running as the standing lens without essentialization of the camp as such or existential stakes. Penetration 4 as below. Elite adoption 4 because a sitting White House adviser voices it as his own and Musk amplifies, but officials beyond that one circle have not taken it up. Lock 3 because the targeted movement responds on the record without the pair yet running on each other as primary evidence.
Mirrors: On the same triggers the opposing carrier set reads the approvals rather than the protests as purchased, and the targeted movement answered this frame on the record, with Kremer denying any Moskovitz money in the same Examiner piece and KSL rally attendees directly rebutting the paid-and-bused claim.
Keeps coming back:
- Humans First New York launch meeting and the Sacks and Musk amplification cycle, March 2026
- Tremonton Stratos approval and the Utah protest wave, May 2026
- Humans First open-letter cycle and the Bull Moose funding report, May 2026
- New York moratorium and the July 18 national protest day, July 2026
Source record: counter-hate-monitor/counter-hate-work/special-reports/2026-tech-ai-fault-line/cards.json#pol-astroturf-psyop
The AI fault line, January to July 2026
Thirteen narratives, every carrier quote, the full scoring, and the coverage notes on what we couldn't verify. This page is the plain-language cut; that one is the whole thing, including the polarization methodology that defines every score here.