"Why Americans dread AI" is an opinion column by Edward Luce, the Financial Times US national editor and columnist, published in the Financial Times on 12 May 2026 (roughly 900 words). It argues that a Silicon Valley narrative — that AI is inevitable and that domestic regulation is either "woke" or a gift to China — has produced political fatalism even as a clear US public majority believes AI does more harm than good. As an opinion column making a political-economy argument rather than reporting new empirical findings, it is treated as a position rather than as evidence.
Summary of argument
Luce's central contention is that the inevitability framing around AI is a public-relations achievement rather than a fact, and that it has discouraged democratic control over the technology despite broad public unease. He applies the label "pluto-populist" to Donald Trump's alliance with the tech industry, describing a combination of Silicon Valley funding and populist political rhetoric directed against safety regulation. The term is presented as Luce's coinage or popularization and is attributed to him on that basis.
Luce also advances a dual-positioning critique of the industry's stance on China: "At home they are hawks. Abroad they are doves. On the one hand, they shower money on those in Washington who say that domestic regulation will help China. On the other hand, they lobby Trump to sell more of their best stuff to China." He cites this as a contradiction with growing weight in the policy debate, pointing to Nvidia's H200 sales to China, Anthropic's refusal of access to the Chinese think tank Mythos, and Jensen Huang joining a Trump China trip.
Against the inevitability narrative, Luce quotes Karen Kornbluh, former director of the National AI Office: "AI is not some deus ex machina. It is humanity's creation and our choice whether to exercise democratic control or not." The quotation functions as a counter-frame to inevitability rhetoric such as the "10× Industrial Revolution" framing associated with Demis Hassabis (Demis Hassabis on our AI future: 'It'll be 10 times bigger than the Industrial Revolution — and maybe 10 times faster' — Steve Rose (The Guardian, August 2025)).
Public-opinion claim
Luce anchors the column in a claim of near-consensus public unease: "Fear of AI is as close to a national consensus as it gets. A clear majority says that AI will do more harm than good. In one recent NBC poll, AI's net negative rating ranked below ICE, the disliked immigration enforcement agency." The NBC poll is cited but not directly examined in the column. The claim parallels the public-opinion case in The AI Backlash Could Get Very Ugly — Lila Shroff (The Atlantic, May 13 2026) and the cross-ideological data that source cites.
Regulatory history
Luce compresses recent US AI-regulation events to support his fatalism thesis:
- Trump "big beautiful bill" (July 2025) — a two-page state-preemption clause was inserted; by Luce's account, "public clamour forced the drafters to drop" it.
- Defense budget (December 2025) — the same state-preemption insertion was attempted and dropped again.
- Trump March 2025 executive order — measures Congress had rejected were enacted by executive order, including provisions Luce characterizes as industry-friendly carve-outs, among them leaving child safety to parents rather than to AI chatbots.
- Post-Mythos preview (April–May 2026) — Trump said AI platforms should seek approval for new models, on a voluntary basis with the Commerce Department as adjudicator. Luce's verdict: "voluntary, however, which makes it meaningless. Moreover, Howard Lutnick's commerce department will be the adjudicator, which is like asking Trump to audit his own taxes." (See AI Pre-Release Vetting.)
Beijing summit anticipation
The column was written on 12 May 2026, ahead of the Trump–Xi summit on 14–15 May. Luce forecasts that Trump may set himself against the foreign-policy consensus by raising AI cooperation with Xi, but cautions that "a Trump overture on AI could be helpful to Silicon Valley. Nvidia's Jensen Huang wants to sell more advanced chips to China." He adds that "the notion that Trump would negotiate global guardrails is as real as a chatbot's hallucination." Dev-log reporting from 14 May (2026 05 14 1402 Ai Developments) records Scott Bessent confirming that AI-guardrails discussions occurred, consistent with Luce's forecast that the topic would arise and with his expectation of thin deliverables.
Provenance and caveats
The column is an FT opinion piece; Luce is making a political-economy argument, not reporting new empirical findings, and it is treated as a position. "Pluto-populist" is Luce's coinage or popularization and is attributed to him. Several claims are characterizations rather than analysis — the "Lutnick = Trump auditing his own taxes" line is rhetorical — and are useful as framing rather than as policy analysis. The NBC poll citation is real but not directly examined in the column.
Tracked claims and their confidence as assessed against the single source:
- NBC poll: AI net negative rating below ICE — medium-high (a specific poll exists; the citation should be tracked).
- Trump preemption clause inserted twice (July 2025 BBB; December 2025 defense budget) and dropped both times after public pressure — high.
- Trump March 2025 EO enacted preemption and industry carve-outs that Congress had rejected — high (verifiable via the EO itself).
- Mythos pre-approval regime is voluntary with Commerce as adjudicator — high as of May 2026; whether it hardens into a mandatory regime over the following six months is unresolved.
- Tech sector lobbies both sides of the China-policy question (hawkish at home, dovish abroad) — Luce's framing; high as a description of revealed behavior.
Relationships
- supports: AI Political Cleavages — pluto-populist framing
- supports: AI Pre-Release Vetting — Luce's argument that voluntary + Commerce-as-adjudicator = meaningless
- supports: State-Level AI Regulation — twice-rejected preemption push, March 2025 EO doing it anyway
- supports: Eo Trump March 2025 State Preemption (or analogous) — primary Luce source on the EO
- supports: AI-Driven Political Violence — public-opinion backdrop to backlash
- related: The AI Backlash Could Get Very Ugly — Lila Shroff (The Atlantic, May 13 2026) — same public-opinion / Bernie-to-Bannon arc
- related: Even Silicon Valley Says That AI Is a Bubble — Lila Shroff (The Atlantic, March 2026) — Luce's "tides" framing of inevitability ≈ Shroff's bubble-defender ideology
- contradicts: Demis Hassabis on our AI future: 'It'll be 10 times bigger than the Industrial Revolution — and maybe 10 times faster' — Steve Rose (The Guardian, August 2025) — Luce treats Hassabis-style inevitability as a PR feat, not a fact