Back to AI & Tech

What worries people about AI

Source reportMethodology

Overview

AI concern is broad across multiple risk areas. Misinformation, bias, and job loss each draw at least 80% at the somewhat-concerned level or higher.

The emotional measure points the same way. About 50% agree that they feel anxious about the rise of AI, compared with 22% who disagree.

Topline

86% are concerned about AI misinformation, 81% about bias, and 82% about job loss.

Misinformation in AI output

  • Extremely concerned 32.6%
  • Somewhat concerned 27.6%
  • Very concerned 25.5%
  • Not at all concerned 9.2%
  • I’m unfamiliar with this topic 5.2%

2025 · base n 1,519 · +/- 3.3%

AI Adoption Survey August 2025

View source data

Misinformation is the highest concern

About 86% are at least somewhat concerned about misinformation in AI output.

The strongest responses are substantial too: 33% are extremely concerned and 26% are very concerned.

Bias and job loss are close behind

Bias in AI model training and output draws about 81% at least somewhat concerned.

AI taking jobs from humans is similarly high at about 82%.

Stacked breakdown

50% agree that they feel anxious about the rise of AI.

I feel anxious about the rise of AI

Strongly agree
16.7%
Somewhat agree
33.0%
Neutral
28.6%
Somewhat disagree
11.7%
Strongly disagree
10.1%

2025 · base n 1,519 · +/- 3.3%

AI Adoption Survey August 2025

View source data

Anxiety is common, but not universal

About 50% agree that they feel anxious about the rise of AI, while 22% disagree and 29% are neutral.

Concern about humans developing relationships with AI is lower than misinformation or jobs but still broad, with about 67% at least somewhat concerned.

Methodology

Full methodology
Mode
Verasight panel recruited via random address-based sampling, random person-to-person text messaging, and dynamic online targeting
Population
United States adults
Field dates
2025-09-03 → 2025-09-08
Base (unweighted)
1,519
Margin of error
+/- 3.3%
Module
AI Adoption Survey August 2025
Sponsor
Verasight
Weight variable
weight
Weighting targets
age, race/ethnicity, sex, income, education, region, metropolitan status

Sources

[2]

Citation

AI Adoption Survey August 2025, fielded September 3-8, 2025, N=1,519 United States adults, +/- 3.3%.

https://reports.verasight.io/r/august-ai-survey-2025#concerns-regarding-ai

Verasight survey methodology

How Verasight conducts surveys.

This page describes the Verasight general survey contract, separate from how the Data Library packages it. When a wave report is published, its field dates, sample sizes, and module breakdown are listed in that report.

Mode
Verasight panel recruited via random address-based sampling, random person-to-person text messaging, and dynamic online targeting.
Population
US adults age 18+.
Sample design
Surveys are run as omnibus or single-topic waves. Omnibus waves are split into modules with their own respondent set, typically around one thousand respondents per module.
Field window
Each wave specifies its own field dates. Most omnibus waves field across roughly two weeks.
Weighting
Per-module weighting to CPS targets including age, race and ethnicity, sex, income, education, region, and metropolitan status.
Partisanship benchmark
Pew Research Center's NPORS benchmarking surveys, three-year running average.
Vote benchmark
2024 presidential vote population benchmarks.
Margin of error
Typically about plus or minus 3.4 to 3.6 percent per module at standard module sizes. Question-level MoE is recomputed when a base shrinks materially below the module baseline.
Reporting
The canonical report hub is reports.verasight.io. Verified legacy reports remain on their original hosts, and unpublished reports are marked pending.
Transparency
Verasight is a member of the American Association for Public Opinion Research Transparency Initiative.

The canonical report hub is reports.verasight.io; older reports may remain on verified legacy hosts.