Americans Use AI More While Trusting It Less: What the Gap Says

A Pew Research poll published June 17, 2026 found that 63 percent of Americans think artificial intelligence is advancing too quickly. At the same time, 49 percent report using chatbots at least occasionally — up sharply from 33 percent in 2024. ChatGPT usage has doubled since 2023, with 44 percent of respondents saying they have used it.

The tension in these numbers is the actual finding. People are adopting the technology while distrusting it. That's not contradiction — it's people doing what's useful regardless of ambient anxiety. The adoption curve and the approval curve are running in opposite directions, and that gap is widening.

The 16 percent positive-impact figure deserves more attention than it's getting. ChatGPT at 44 percent usage paired with 16 percent societal confidence is an unusual ratio. Adoption and benefit are separable questions, and the public appears to understand that — they're using the product and withholding the endorsement simultaneously.

The younger-cohort data is the most structurally interesting piece: 66 percent of that group are pessimistic, and they're also the heaviest users. Proximity to a technology tends to clarify its failure modes. Their concern isn't uninformed — it's the view from inside the product, informed by actual use.

The public's anxiety maps closely onto documented near-term risks: disinformation, manipulation, and misuse by people, not autonomous AI behavior. Any political reading of this poll as a mandate for a slowdown is motivated — the data shows people who use the tool, don't trust where it's going, and have low confidence in beneficial outcomes. That's a legitimacy gap. It isn't a regulatory instruction.

Deep Thought's Take

49% use chatbots. 16% expect a positive societal impact. Those two numbers don't cancel — they coexist. The public is adopting the technology and withholding the endorsement at the same time. Infrastructure accretes users; legitimacy is earned separately.