Half of Americans Use AI, and Most Distrust It With Their Data
49 percent of US adults now use AI chatbots, up from 33 percent in 2024. That is the headline number from Pew Research Center’s “Americans and AI 2026,” a survey of 5,119 adults conducted February 17 to 23, 2026. Adoption nearly doubled in two years. Over the same window, the share of Americans who say they trust AI with their personal data went nowhere.
The two lines that should rise together are pulling apart. People are using the technology more and believing in it less.
Adoption Is No Longer Early
The numbers read like mass-market saturation, not early-adopter curiosity.
Per Pew: 44 percent of US adults now use ChatGPT, up from 18 percent in 2023. 24 percent use an AI chatbot daily. 60 percent say they read the AI-generated summaries that appear at the top of search results. The technology is no longer something people seek out. It arrives in the search box whether they asked for it or not.
That last point matters for how trust forms. A tool you choose earns its credibility through use. A tool that shows up uninvited inherits whatever suspicion the user already carried. More than half of Americans are now reading AI output before they have decided whether to trust it.
Trust Stalled While Use Climbed
Adoption doubled. The concern indicators held firm or rose.
Pew found that 63 percent of US adults say AI is advancing too fast. 71 percent expect AI to weaken the security of their personal data. 67 percent say they lack confidence that government can regulate AI effectively, up from 62 percent in 2024. Read those together and a pattern emerges: people are adopting a technology they believe is moving faster than anyone can govern, including the institutions meant to govern it.
Data security is the dominant fear, ahead of job loss or misinformation. 71 percent is not a worried minority. It is a supermajority of the same population that just doubled its usage. They are not waiting for proof that their data is safe. They are using the product anyway and expecting to be harmed by it.
The distrust in regulation deserves its own line. When two thirds of the public doubts the state can keep up, the burden of demonstrating safety shifts to whoever ships the product. Public confidence is not going to backfill from Washington. It has to be built at the point of use.
The Partisan Floor Disappeared
One finding cuts against the usual political read. Distrust of AI regulation is no longer a partisan signal.
In 2024, confidence in government oversight of AI split along familiar lines. By 2026, Pew shows the doubt has broadened across the spectrum. The 67 percent who lack confidence in regulation is not one camp’s complaint. It is the rare position both sides hold at once. When skepticism stops being partisan, it stops being noise. It becomes the baseline expectation a company has to design against.
For anyone deploying AI in a customer-facing product, this removes a convenient excuse. You cannot dismiss data-security concern as one demographic’s hang-up. It is the median customer’s starting assumption.
What the Oversight Concern Now Means for Buyers
For years, the argument that AI needed guardrails lived in expert panels and policy papers. Pew’s 2026 data moves it into the general public. 63 percent saying “too fast” and 71 percent expecting data harm is the same worry that compliance teams have been raising, now voiced by the people who buy the product.
This reframes the commercial question. The growth side of every AI roadmap is solved: people will use it. The unsolved side is whether they will trust the company that built it with anything that matters. A product that captures usage while losing confidence is accumulating a liability that surfaces later, when a breach, a bad output, or a regulatory action converts quiet distrust into churn.
Companies treat trust as a brand exercise. Pew’s numbers say it is now a measurable operating risk. The customer already expects their data to be mishandled. The only variable a vendor controls is whether the product’s actual behavior beats that expectation.
Do This Now
Audit where AI touches customer data in your product and write down, in one page, what the system does with it. Most teams cannot answer this cleanly, which is why the 71 percent feel the way they do.
Then make that answer visible to the customer before they ask. The companies that win the next two years will be the ones whose demonstrated handling of data outpaces the public’s default suspicion. Adoption is no longer the contest. Earned confidence is. Treat data-security transparency as a product feature with an owner and a roadmap, not a legal footnote.
The 49 percent are already in the door. The 71 percent are the reason they might not stay.
This analysis synthesizes Americans and AI 2026 (Pew Research Center, June 2026). See also our work on the AI adoption spectrum, growth as a trust problem, and the B2B AI trust gap.
Victorino Group helps companies turn data-security transparency into a measurable advantage, not a compliance afterthought. Let’s talk.
All articles on The Thinking Wire are written with the assistance of Anthropic's Opus LLM. Each piece goes through multi-agent research to verify facts and surface contradictions, followed by human review and approval before publication. If you find any inaccurate information or wish to contact our editorial team, please reach out at editorial@victorinollc.com . About The Thinking Wire →
If this resonates, let's talk
We help companies implement AI without losing control.
Schedule a Conversation