AI was supposed to win people over by now — it hasn’t
AI's public reputation is deteriorating despite the technology's rapid advancement, with growing evidence that consumer trust and enthusiasm are declining rather than building over time. Recent polling shows rising unease about AI's role in daily life, distrust of industry leaders, and concern that development is moving too fast, while tech companies are also facing political backlash over data centre construction and are having to sweeten local deals to offset opposition. This matters because it undercuts the assumption, common among AI companies that have raised hundreds of billions of dollars, that widespread adoption would naturally translate into public acceptance.
Pew Research found 52% of Americans are now "more concerned than excited" about AI's growing presence in daily life, up from 37% in 2021, while a May Economist/YouGov poll found over 70% believe AI is advancing too quickly. A CNBC poll of 18- to 34-year-olds found most don't trust prominent AI industry leaders to act responsibly, and the Wall Street Journal reported tech firms offering job guarantees, clean water investment and even $50,000 teacher bonuses in Louisiana to smooth opposition to new data centres. The backlash appears sharper than that faced by earlier transformative technologies such as the iPhone or the internet, and is coinciding with a resurgence of interest among young people in "retro" alternatives — dumbphones, film cameras, cassette tapes and analogue hobbies like knitting and jigsaw puzzles — alongside a shift towards in-person activities over online alternatives.
- Public trust in AI is falling, not rising, despite rapid progress
- Polls show growing concern, distrust of AI leaders, and unease
- Backlash is fuelling a "retro tech" and offline-hobbies trend among young people
New here? Start with this
AI companies have poured hundreds of billions of pounds into building artificial intelligence tools, expecting that as people used them more, trust and enthusiasm would follow naturally. New survey evidence suggests the opposite is happening: even as the technology improves, more people say they feel uneasy about it rather than reassured.
Several polls illustrate this shift. In the United States, the proportion of people who say they are more concerned than excited about AI has grown sharply since 2021, most now believe the technology is being developed too quickly, and younger adults in particular say they don't trust the leaders of major AI firms to act responsibly. Separately, tech companies building the data centres that power AI have faced local opposition strong enough that some have offered financial incentives, such as job guarantees or community payments, to win support.
This matters because public trust underpins how quickly and widely new technology gets adopted, and because the pattern differs from earlier breakthroughs like the smartphone or the internet, which faced less resistance at a similar stage. The unease has also coincided with renewed interest among some young people in simpler, non-digital alternatives, from basic phones to analogue hobbies, suggesting the public response to AI may be more complicated than the industry anticipated.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Growing public wariness about AI reflects a reasonable, evidence-based response rather than mere technophobia. People are entitled to weigh the concrete costs they can already see – job insecurity, strained local water and power resources from data centres, and a demonstrated willingness among some AI leaders to prioritise speed over safety – against benefits that remain largely promised rather than delivered. Scepticism about placing enormous trust in a small number of unaccountable companies, and a preference for slowing down until governance catches up with capability, is a prudent stance rather than an irrational one, and the turn toward analogue hobbies and in-person life can be read as a healthy reassertion of autonomy rather than mere nostalgia.
The case against
The current backlash can equally be seen as a normal, even necessary, phase that most disruptive technologies pass through before their value becomes fully apparent, and it would be a mistake to read short-term polling dips as a verdict on AI's long-term trajectory. Advocates of continued rapid development argue that unfamiliarity and sensationalised coverage of worst-case scenarios naturally outpace public understanding of everyday benefits such as medical research, productivity gains and scientific discovery, and that similar unease accompanied electricity, the internet and smartphones before they became indispensable. On this view, companies offering community investment around data centres are reasonably addressing legitimate local concerns rather than admitting failure, and dismissing today's caution as fear rather than foresight would be to prejudge a debate that deserves patience and better public communication rather than a slowdown in innovation.