About half of Americans want tight limits on government use of AI
Core finding
Verasight surveyed 1,000 U.S. adults from September 25 to 30, 2026. 52.2% of Americans want the federal government's use of AI tightly restricted, agreeing it makes too many mistakes and removes human judgment. 24.6% would use it wherever it improves efficiency, and 23.2% lean neither way.
Respondents weighed two statements. Using AI wherever it improves efficiency draws 7.2% strong and 17.5% some agreement.
Topline distribution
About half of Americans want tight limits on government use of AI.
Artificial Intelligence (AI) is designed to learn tasks that humans typically do, such as recognizing speech, analyzing data, or writing text. Thinking about how the federal government uses AI as part of its official work, which comes closer to your view, even if neither is exactly right? Statement A: AI can make government more efficient, so officials should use it wherever it helps. Statement B: AI makes too many mistakes and removes human judgment, so its use should be tightly restricted.
n = 1,000 · Sep 25–30, 2026 · MoE ±3.4%
Methodology
Full methodology →| Mode | Verasight panel recruited via random address-based sampling, random person-to-person text messaging, and dynamic online targeting |
|---|---|
| Population | US adults age 18+ |
| Field dates | Sep 25–30, 2026 |
| Base (unweighted) | 1,000 |
| Margin of error | ±3.4% |
| Module | Technology, Security, and Global Affairs |
| Sponsor | Verasight |
| Weight variable | weight |
| Weighting targets | age, race/ethnicity, sex, income, education, region, metropolitan status |
Sources
- reports.verasight.io/r/2026-153-apsa-omnibus-report
Statement B: AI makes too many mistakes and removes human judgment, so its use should be tightly restricted.
Cite this question
APSA Omnibus Survey #2026-153, fielded September 25-30, 2026, N=1,000 US adults age 18+, +/- 3.4%.