"The AI Backlash Could Get Very Ugly" is a May 13, 2026 essay in The Atlantic by technology writer Lila Shroff. It argues that opposition to AI, already strong in polling, is structurally primed to intensify as data-center opposition spreads, cross-ideological political messaging gains traction, and labor displacement materializes. The essay pairs with Shroff's earlier "Even Silicon Valley Says That AI Is a Bubble" (March 2026); both link bubble dynamics to social and political fallout. It is an opinion essay synthesizing existing polling, threat reporting, and academic work into a prospective argument rather than original empirical research.
Summary of argument
Shroff argues the anti-AI backlash is positioned to grow, organized around three vectors.
The first is bipartisan public opposition that is already present. The essay states the United States ranks among the countries most concerned about AI, cites an NBC poll in which AI's net negative rating fell below that of ICE, and reports that a majority of Americans believe AI does more harm than good. It notes that both Democrats and Republicans oppose new data centers.
The second is local data-center opposition producing both legislative outcomes and acts of violence. Maine passed the country's first statewide data-center moratorium, which was vetoed by Governor Mills. Q1 2026 saw a record number of canceled data-center projects following local pushback, which Shroff attributes to Heatmap News reporting. April 2026 brought the Altman Molotov attack and the Indianapolis "No Data Centers" shooting; Shroff treats these as early signals rather than aberrations. The essay also draws on Soufan Center reporting of a rise in "direct threats" against individuals, policymakers, and corporations involved with AI, noting that the most common online threats involve "physical sabotage of proposed or operational data centers."
The third is a cross-ideological political coalition Shroff labels "Bernie-to-Bannon," spanning figures who do not align on the standard left-right axis. Bernie Sanders is quoted: "AI oligarchs do not want to just replace specific jobs. They want to replace workers." Steve Bannon is quoted saying Silicon Valley "does not care about the little guy," on a podcast titled Stopping the AI Oligarchs From Stealing Humanity. Josh Hawley (R-MO) is quoted: "I have no doubt that these companies are going to get filthy rich, but is it going to be good for children?...Is it going to be good for the American worker?" Shroff reports that Blue Rose Research has found populist anti-AI messaging is electorally effective for Democrats heading into the midterms. The essay also quotes Mark Warner's concern that "populism from both the left and the right" could curb innovation.
The essay explicitly extends the framework of Veilleux-Lepage's CTC paper to a mainstream audience, citing the paper by name and quoting its language describing local policymakers as proxies for "structural anger" when national figures are unreachable. Shroff also uses an Industrial Revolution analogy in the opposite direction from how Silicon Valley typically invokes it: where the industry cites the eventual productivity payoff, Shroff notes that "living through it was another matter entirely…attacks on the industrialists themselves" during the actual transition.
Income-divided optimism
Shroff cites a Quinnipiac poll finding that the group of Americans most optimistic about AI in their daily lives are those making more than $200,000. The essay frames this as relevant to both political coalition formation and Veilleux-Lepage's economic-order grievance domain.
Industry response
Shroff notes that the AI industry is beginning to respond to the backlash. a16z published an essay declaring "job apocalypse" to be "baseless fantasy," framing AI labor disruption as overblown. Sam Altman shifted from a 2023 statement that "jobs are definitely going to go away, full stop" to a May 2026 position that "Jobs doomerism is likely long-term wrong." The essay also describes discussion on Tech-Twitter of cosmetic remedies, including that data centers should be "beautiful" and that better messaging campaigns are needed. Shroff's response: "If the tech industry truly believes that a simple change in messaging will quell the backlash, then they are misunderstanding the problem entirely."
The essay's central claim that the backlash will intensify is prospective and normative; its accuracy depends on whether AI-driven displacement materializes at the scale the bear case anticipates (see AI Labor Disruption).
Key quotes
"AI generates the structural conditions historically associated with the onset of political violence." — Veilleux-Lepage, quoted by Shroff
"Where national figures are unreachable, local policymakers who approved the data center become the proxies for the same structural anger." — Veilleux-Lepage via Shroff
"Disruption has winners and losers. For many Americans, they're not convinced they're going to be the winners, and they base that conclusion on the history of technology over the last 20 years." — Nathaniel Persily, Stanford
Provenance and caveats
The piece is an Atlantic essay, not original empirical research; Shroff synthesizes existing data (polls, Soufan threat reporting, the Veilleux-Lepage paper) into a prospective argument. The "Bernie-to-Bannon coalition" is rhetorical framing rather than a formal coalition, and whether it produces a coherent legislative agenda is uncertain.
Confidence in specific tracked claims varies. The Q1 2026 record of canceled data-center projects after local pushback (Heatmap News reporting cited by Shroff) and the characterization of the Maine moratorium as the country's first statewide data-center moratorium (vetoed) are held at high confidence. The Bernie-Bannon-Hawley convergence on anti-AI rhetoric is high confidence, with quotes verifiable. The Quinnipiac finding that the cohort most optimistic about AI is households earning more than $200,000 is held at medium-high confidence as a single poll consistent with broader survey patterns. The Soufan Center reporting of a rise in direct threats against AI individuals, policymakers, and corporations is held at medium confidence, as the primary Soufan reporting was not directly examined.
Relationships
- supports: AI-Driven Political Violence — primary mainstream-media bridge from academic framework
- supports: AI Political Cleavages — Bernie-to-Bannon coalition
- supports: AI Labor Disruption — backlash dynamics; income-divided optimism
- supports: Data Center Siting / AI Power Politics — Maine moratorium, Q1 cancellations, Indianapolis pushback
- supports: Beyond Misuse: Artificial Intelligence, Grievance, and the Future Landscape of Political Violence — Yannick Veilleux-Lepage (Combating Terrorism Center at West Point, April 2026) — cites and extends the CTC framework to mass audience
- related: Even Silicon Valley Says That AI Is a Bubble — Lila Shroff (The Atlantic, March 2026) — same author's bubble essay; bubble + backlash are linked theses
- related: Sam Altman — Altman Molotov attack as proximate trigger; Altman's pivot on jobs doomerism