America Uses AI But Doesn't Trust It

The Numbers Are Worse Than You Think

The NBC News poll, conducted February 27 through March 3, 2026, surveyed 1,000 registered voters via a mix of telephone interviews and text-based online responses. It carries a margin of error of plus or minus 3.1 percentage points. The results, compiled by Democratic polling firm Hart Research Associates and Republican firm Public Opinion Strategies, are striking in their consistency across methodology.

Just 26% of registered voters reported a positive view of artificial intelligence. 46% reported a negative view — a net favorability of minus 20. For context, according to the same NBC News poll:

• President Donald Trump: 41% positive / 53% negative (net: -12)
• Immigration and Customs Enforcement: 38% positive / 56% negative (net: -18)
• The Republican Party: 37% positive / 51% negative (net: -14)
• AI: 26% positive / 46% negative (net: -20)
• The Democratic Party: 30% positive / 52% negative (net: -22)
• Iran: 8% positive / 61% negative (net: -53)

AI sits third from the bottom in a survey that included the Iranian government and a deeply unpopular American political party. The only figures above water in the poll were Pope Leo XIV (net +34) and comedian Stephen Colbert (net +10).

Beyond favorability, the poll found that 57% of registered voters believe the risks of AI outweigh its benefits, versus 34% who said the opposite. Asked which political party handles AI better, 20% said Republicans, 19% said Democrats, 33% said neither, and 24% said they were about the same. Republican pollster Bill McInturff of Public Opinion Strategies told NBC News the findings suggest AI is "up for grabs" as a political issue — meaning neither party has successfully claimed it as an asset.

The Paradox: They Use It Anyway

Here is where the data gets complicated. The same NBC News poll found that 56% of voters said they had used an AI platform such as ChatGPT or Microsoft Copilot within the previous two months. A separate Brookings Institution survey, also cited in reporting on the poll, found a similar figure for general AI tool usage among American adults.

Pew Research Center's March 12, 2026 analysis — drawing on five years of Center surveys — adds further texture. As of a June 2025 Pew survey, 31% of American adults said they interact with AI at least several times per day, up from 22% in February 2024. 64% of U.S. teenagers aged 13 to 17 said they ever use an AI chatbot, according to a fall 2025 Pew survey of that age group.

Yet Pew also found that 50% of U.S. adults said the increased use of AI in daily life makes them feel more concerned than excited, per the June 2025 survey. Just 10% said they were more excited than concerned. That 50% figure is up sharply from 37% in 2021, when Pew first asked the question.

The pattern is unmistakable: Americans are using AI at higher rates while simultaneously growing more worried about it. A February 2026 poll by research firm Verasight found 56% remain anxious about AI's rise. The Verasight authors wrote that "until companies and institutions building AI systems can address fundamental concerns about trustworthiness, privacy, and social responsibility, Americans will remain trapped between enthusiasm and anxiety, using AI while quietly wondering if they should."

The Trust Problem Is Specific and Severe

The credibility gap isn't just a vague cultural discomfort — there are hard numbers on it. An October 2025 survey by OnMessage Inc., released during a panel at NAB Show New York by the National Association of Broadcasters, found that only 26% of consumers trust information produced by AI, while 68% said AI-produced information is not trustworthy. The same survey found 76% of respondents were concerned about AI "stealing or reproducing journalism and local news stories," with 51% describing themselves as very concerned.

Pew's analysis found that about half of U.S. adults said AI will worsen people's ability to think creatively and form meaningful relationships with others. Views were more optimistic for narrow technical applications — Americans were more likely to see AI having a positive impact on medical care (44% positive vs. 19% negative) than on education (24% positive) or how people do their jobs (23% positive).

The NBC News poll also found deep partisan and demographic fault lines. Among voters aged 18 to 34, the net favorability rating for AI was minus 44. Among women aged 18 to 49, it was minus 41. The two groups with the most favorable views of AI were men over 50 (net +2) and upper-income voters (net +2). Republicans were evenly split — 33% positive, 33% negative. Independent voters were 26% positive and 48% negative. Democrats: 20% positive, 56% negative.

What's Driving the Fear: Jobs

The NBC News poll comes against the backdrop of mounting evidence — and mounting anxiety — about AI's effect on employment. The poll noted that the job market contracted in five of the nine months preceding its publication, and that 25% of unemployed workers held four-year college degrees in November 2025, described in NBC's reporting as a record figure.

A Stanford University research team found, according to the NBC News report on the poll, that workers aged 22 to 25 in industries with the most exposure to generative AI experienced a 16% relative decline in employment since late 2022.

Anthropic — the company that built Claude — released its own labor market analysis in early March 2026. Drawing on U.S. Census Bureau occupational data, Anthropic researchers found that workers in the professions most exposed to AI automation are disproportionately "older, female, more educated, and higher-paid," according to the report as cited by CBS News, Business Insider, and Euronews. The occupations Anthropic identified as having the highest observed AI exposure include business, finance, law, and coding roles.

Anthropic CEO Dario Amodei had warned as early as 2025 that AI could disrupt half of entry-level white-collar work. Fortune characterized Anthropic's March 2026 report as suggesting a possible "Great Recession for white-collar workers."

Republican pollster Micah Roberts of Public Opinion Strategies told NBC News: "Voters' economic circumstances and education level are also strongly dividing attitudes. You can see clearly in this data... how negativity toward AI drops as people become higher-educated." Roberts also noted that concerns were "especially relevant to younger voters and women under 50" — precisely the groups the Anthropic data identifies as most at risk.

A Policy Void Neither Party Can Fill

The NBC News poll offers an unusual finding about the political landscape: AI is the one issue polled on which neither party holds an advantage over the other. The 20% who said Republicans handle AI better and the 19% who said Democrats do are statistically indistinguishable. The dominant answer — 33% — was that neither party handles it well.

The Trump administration has made clearing regulatory obstacles for AI development a stated priority. In an interview with NBC News, President Trump pushed back against job-displacement fears, saying: "They said the internet was gonna do — everything was gonna do — robots are gonna kill jobs. Everything's gonna kill jobs. And you end up, if you're smart, doing great."

That message has not moved the needle. The demographic groups most hostile to AI — young workers and women under 50 — are precisely those whose economic concerns the administration has struggled to address. Big tech companies have collectively committed up to $700 billion in AI capital expenditures, according to Fortune's reporting on the poll — investments that rival the annual GDP of mid-sized developed economies. The data center construction boom tied to that spending has also generated electricity bill increases for consumers, adding another source of concrete grievance on top of abstract fears.

The Bottom Line

The convergence of data from NBC News, Pew Research Center, the National Association of Broadcasters, Anthropic, and Verasight tells a coherent and uncomfortable story: artificial intelligence is being adopted at a historically rapid pace by a public that neither trusts it nor wants it to proceed unchecked.

That split — between what Americans do and what Americans fear — may be the most politically consequential technology story of 2026. AI is already in workplaces, classrooms, and phones. The public's skepticism hasn't slowed that deployment. But the polling suggests that when AI becomes a campaign issue — and it likely will — neither party starts with an advantage, and fear runs deep across ideological lines.

The Verasight poll framed the moment precisely: a country using a technology it doesn't fully trust, "quietly wondering if they should." That's not a verdict on AI. It's a description of where the conversation starts.