AI's approval problem now has a price: job guarantees, water money, teacher bonuses

Pew Research now puts the share of Americans who say they are "more concerned than excited" about the increased use of AI in daily life at 52%, up from 37% in 2021 ^[1]. In the same week, Axios reported that the National Republican Senatorial Committee had sent a memo to top AI companies warning that US data centres are hurting the party's chances in a key Ohio election ^[2], and the Wall Street Journal reported that tech companies facing a public relations crisis over their data centre plans are sweetening deals with job guarantees, clean water investments and other local perks ^[3].

That is the part operators should care about. The Pew shift is 15 points, roughly a 41% increase on the 2021 base ^[4], and it is no longer just a survey artefact: it is showing up as concessions attached to permits. In one Louisiana parish, according to the Journal's reporting, the package included $50,000 bonuses for teachers ^[5]. Whatever you call that internally, it is a cost of goods for compute capacity, and it is set by people who do not work at your company.

The rest of the polling stack points the same way. A CNBC poll of 18- to 34-year-olds presented respondents with the names of nine top AI industry leaders, and a majority said they do not trust those people to "act responsibly" on AI ^[6]. A May Economist/YouGov poll found more than 70% of Americans think AI is advancing too quickly ^[7]. TechCrunch's read is that AI is drawing more consumer backlash than the iPhone, the personal computer or the internet did at comparable stages of adoption ^[8].

The comfortable internal explanation is that this is a narrative failure. The less comfortable one, which TechCrunch argues, is that consumers understand the trade-off well enough and have priced it: they do not see how AI improves their lives, while being asked to absorb its costs ^[9]. What they mostly encounter is AI features arriving in email, search and televisions whether or not they were requested, AI as a cheating tool in schools, and models trained on other people's intellectual property ^[10]. Airbnb CEO Brian Chesky said on a recent podcast that the backlash is real and is tied to the industry not shipping products that regular people like: "part of it's a narrative issue that we're not talking about AI correctly," he said, "but part of it is we need to actually be developing more products that just regular people can use" ^[11][12].

There is a demand-side tell as well. TechCrunch notes young consumers moving toward dumbphones, point-and-shoot cameras, tape decks and CD players, with AI-free classic iPods fetching high prices on eBay ^[13], alongside a revival of quilting, knitting, puzzles and card games, and in-person run clubs winning attention over online dating apps ^[14]. None of that is a referendum on model quality. It is substitution away from the category.

Three things to watch. Whether the data centre concessions are one-time or recurring, because a recurring local payment changes the unit economics of capacity rather than the marketing budget. Whether the Ohio pattern the NRSC flagged ^[2] repeats in other siting fights, which would make politics a scheduling risk. And whether the next Pew wave moves again, because an industry that raised hundreds of billions on the premise of inevitability ^[15] has no line item for a hostile electorate.