More Americans Fear AI Harming Elections Than Loved Ones Dying

By James Calloway

51.9% of Americans say they are afraid or very afraid that artificial intelligence will harm the political process — a fear that now ranks third in the country, ahead of loved ones becoming seriously ill (51.8%) or dying (50.6%). That finding comes from Chapman University's 12th annual Survey of American Fears for 2026, which asked about AI for the first time in the project's history.

The private California research university commissioned NORC at the University of Chicago to survey 1,221 adults online this past April. Two AI-related fears landed in the top 10: harm to the political process (No. 3) and harm to the economy (49.7%, No. 7). With midterm elections approaching, the poll captures a public already bracing for AI's role in campaigns and voting.

The stakes are straightforward. If half the electorate distrusts the information environment around an election, that distrust itself becomes a political force — regardless of how much AI-driven interference actually occurs. The survey measures perception, not verified harm, and Chapman does not claim otherwise.

What topped the 2026 fear list?

Corrupt government officials remained the No. 1 fear for the second straight year, at 61.1%. The prospect of the US entering another world war jumped from No. 6 in 2025 to No. 2, at 58%.

The full top 10, as measured by Chapman:

• Corrupt government officials — 61.1% (No. 1)
• US involvement in another world war — 58% (up from No. 6)
• AI harming the political process — 51.9% (new entry)
• People I love becoming seriously ill — 51.8% (No. 2 in 2025)
• People I love dying — 50.6% (No. 5 in 2025)
• Economic/financial collapse — 50.6% (No. 3 in 2025)
• AI hurting the economy — 49.7% (new entry)
• Government tracking of personal data — 48.9% (No. 10 in 2025)
• Biological warfare — 48.2% (No. 12 in 2025)
• Terrorist attack — 47.3% (No. 15 in 2025)

Read plainly: Americans newly rank two AI scenarios alongside war, illness and economic collapse. That is a remarkable shift for a category the survey had never measured before.

How deep does AI anxiety run?

Beyond the top 10, the survey shows AI worry spread across several distinct scenarios, at sharply different intensities:

• AI hurting the environment — 43.2%
• AI harming personal relationships — 42%
• AI replacing people in the workforce — 42%
• Having your work evaluated by AI — 19.3%
• Reporting to an AI manager at work — 19.1%
• Working with AI-enabled technologies — 17.2%

The gap is worth noting. Fears about AI's broad societal effects cluster in the 40s, while fears about AI in one's own daily work sit below 20%. Americans appear far more worried about what AI does to society than about what it does to their job — at least when asked in the abstract.

Other technology fears also registered: cyberterrorism at 45.7%, government use of drones within the US at 34.2%, "technology that I don't understand" at 23.5%, and technological change at work at 12.6%. In short, AI did not displace older tech anxieties; it joined them.

Is the picture entirely negative?

Not according to Chapman's own researchers. The survey also recorded hope about the technology, said Seth Benzell, assistant professor of business and economics at Chapman.

"When you frame it as AI or robots that are going to allow you to build something that you weren't able to do before … those get people a lot more excited," Benzell said.

The caveat follows the promise: framing determines the answer. A survey that emphasizes disruption will find fear; one that emphasizes capability will find enthusiasm. Chapman's own data shows both, depending on the question.

What happens next?

The midterms will be the first national test of whether these fears translate into demands for AI regulation in political advertising, campaign communications and election administration. Expect lawmakers and platforms to cite this 51.9% figure often — and expect it to measure anxiety, not evidence.