Model Republic traces the September anti-'AI safety' online campaign to Innovation Council Action, a $100M pro-AI dark-money group
On Sept 25, 2026 Tyler Johnston (The Midas Project) published an analysis in Model Republic arguing that the wave of attacks on AI-safety advocates in September, which called safety concerns a "psyop" or a "cult", was coordinated by Innovation Council Action, a nonprofit led by former White House deputy chief of staff Taylor Budowich that plans to spend at least $100M on the 2026 midterms and does not disclose its donors.
Key facts
- Innovation Council Action launched in March 2026 to back the Trump administration's AI agenda; plans ≥$100M; led by Taylor Budowich and backed by David Sacks; donors undisclosed (Anadolu, Popular Information, Hoodline)
- Johnston's count: ICA's X posts rose from 45 in July to 453 in Sept 1–24, more than half attacking AI safety; ICA also ran produced ads such as 'Effective Altruism is a cult'
- Senior policy adviser Jordan Schachtel posted 192 times on Sept 8–23, 104 of them hitting at least one anti-safety talking point
- Johnston links the campaign to tabloid stories citing anonymous White House sources and to Trump's Sept 14 post calling AI-extinction fears a 'HOAX'
- Funding by a16z or Nvidia is Johnston's speculation, not established; the analysis is by an AI-safety advocate (The Midas Project)
What happened
After Dario Amodei's "pace the frontier" essay and Hinton's congressional briefing in mid-September, pro-industry accounts pushed a narrative that AI-safety concern was an orchestrated "psyop" by effective altruists. Johnston counted posts and talking points from Innovation Council Action and accounts around it and argued that the backlash was itself a coordinated, well-funded campaign.
Why it matters
AI policy has become a midterm-election issue with nine-figure spending on both sides. Groups that do not have to disclose donors make it hard to see who is funding the fight over AI regulation. The specific attributions in the analysis come from an advocate and are not independently confirmed.
Changelog
- 2026-10-03: created (lead from Zvi, Oct 2)
Related events
- NYT: OpenAI- and Anthropic-aligned super PACs have spent $55.7M on the US midterms, and none of their 95 ads mention data centers ★★★
- Trump says he will form an 'AI Force' and name an AI czar, while calling AI-safety fears a 'hoax' ★★★
- Dario Amodei publishes "We Must Pace the Frontier", calling for a deliberate slowdown ★★★★
- Quinnipiac poll: 77% of Americans favor slowing (47%) or stopping (30%) powerful AI development; 73% fear AI could threaten human survival ★★★
Sources (3)
- discussionModel Republic (Tyler Johnston): Who is behind the AI safety backlash?
- pressAnadolu Agency: Pro-AI group to spend $100M in 2026 US midterms backing Trump tech agenda
- pressPopular Information: AI industry taps January 6 operative to run $100 million campaign
id: 2026-09-25-model-republic-ai-safety-backlash-innovation-council · updated 2026-10-03 · open in the interactive timeline