Two-thirds of Americans think AI is advancing too quickly
TL;DR
Pew Research finds that 49 percent of US adults use chatbots at least occasionally, up from 33 percent in 2024. Adoption is rising, but sentiment is not: 63 percent say AI is advancing too quickly, and only 16 percent expect a positive impact on society. ChatGPT is the most visible product in the data: 44 percent of respondents say they have used it, roughly double the 2023 level.
Nauti's Take
This is not a simple anti-AI backlash; it is an adoption problem hiding in plain sight. People try chatbots because the utility is obvious, then distrust the broader trend because the costs are obvious too: errors, job anxiety, data questions, and speed without enough democratic friction.
Selling AI as destiny will keep backfiring; durable trust comes from boring, useful, verifiable systems.
Briefingshow
The poll captures the core tension in the AI cycle: usage is climbing faster than trust. When people rely on chatbots while expecting social harm, adoption becomes fragile. Companies cannot fix that with bigger launches alone; they need reliability, clear boundaries, and accountability users can actually see.
Sources