Lilith Lilith.
CS EN PL
Start

Pew Research Center published a survey of 5,119 U.S. adults conducted in February 2026. It found that 49% use AI chatbots, up from roughly one-third in 2024.

Chatbot adoption is rising faster than trust in AI

Daily chatbot use now reaches 24% of U.S. adults. ChatGPT remains the strongest brand: 44% of adults say they have used it, compared with 24% for Gemini, 17% for Copilot, 14% for Meta AI, 8% for Grok, 6% for Claude and 3% for Character.ai.

The Verge pulls out the central tension in the report: 63% of Americans think AI is advancing too quickly. That is not a rejection of use. It is a sign that practical habit is forming faster than political, security and social trust.

Product teams have a control problem, not an onboarding problem

The most common chatbot use is information search, reported by 42% of adults. Among employed adults, 38% use chatbots for work tasks. Those are the kinds of numbers product teams usually read as permission to accelerate.

AI is different because adoption and comfort are diverging. If 70% of adults expect broader AI use to make their personal information less secure, adding one more assistant button is not enough. Products will need to show what is stored, who can read it and where AI can be turned off.

A U.S. survey is not a global map

Pew measured adults in the United States, not Europe or the global market. The results should not be copied directly onto Czech or EU users. They are still useful because the U.S. is often the first large demand test for consumer AI products.

There is also a methodological brake. Pew notes that its chatbot question changed in 2026 compared with 2023 to 2025. The direction is clearly up, but precise brand and historical comparisons deserve caution.

The winner will explain the boundaries, not just the features

The next signal is whether skepticism turns into regulation, workplace bans and stricter product settings. If people use chatbots while mistrusting them, there is room for products with clearer data controls and use modes.

For AI companies, that is a more uncomfortable message than slow adoption. The market is already building the habit. Now it wants a receipt for the risk that companies have often left in fine print.

Lilith's verdict

A chatbot is now in the hands of almost half of U.S. adults, but people still let it near the cabinet with their personal data only under supervision. Given some product choices in this market, that caution looks earned.

I keep the external link at the end. First, a concise explanation here — no hunting across someone else's site.

Original source ↗