Anthropic Public Record, wave 1: how about 52,000 Americans see AI
Anthropic AI Attitude Surveys
Anthropic Public Record, wave 1: how about 52,000 Americans see AI
- Published: 2026-06-12
- Sample: 51,993 Americans (age 16+, all 50 states, D.C. and Puerto Rico)
Method and sample
- The first wave of a new, recurring survey series.
- Fielded November 1 to December 11, 2025, online via YouGov.
- 51,993 respondents, weighted to US Census benchmarks.
- State samples range from 232 (Alaska) to 1,902 (New York).
- National margin of error is plus or minus 0.6 points; state-level margins range from 2.6 to 9.1 points.
Top hopes (share ranking it in their top three)
- Curing diseases like cancer or Alzheimer's: 48%
- Helping people with disabilities: 36%
- Making technological progress: 23%
- Making life easier in general: 23%
Top fears
- AI-induced job loss: 64%, the top fear in every state.
- Cognitive dependency: 56%
- Misinformation: 52%
Patterns in job-loss worry
- Top fear among Democrats (67%) and Republicans (62%).
- By state, from Iowa (71%) to Mississippi (57%).
- Postgraduates are nearly 10 points more worried than those with a high school education or less.
- Daily AI users at work: 54%, versus 70% among non-users.
Perceived ability and dependency
- 75% said AI is as good or better than humans at research; 44% in service and support.
- The more competent AI seems in a domain, the more willing people are to use it.
- Dependency worry: 46% among daily users at work, 62% among those who never use it.
- Educators are 2.5 to 3 times more likely than average to report seeing cognitive atrophy firsthand.
What Americans want from government
- 71% said government should be involved in developing and regulating AI.
- Democrats 79%, Republicans 68%, Independents 69%.
- A majority in every state and territory, from D.C. (81%) to Hawaii (63%).
- Priority areas: privacy 56%, child safety 52%, liability for harm 49%.
The trust deficit
- Top actions: hold AI companies legally liable for harm (47%) and prioritize safety over growth (44%).
- Only 15% trust AI companies to decide how AI is developed and used.
- Federal government 20%, state and local 19%, international bodies 20%, independent experts 43%.
- AI companies scored lowest of every institution tested.
The heaviest users
- About 6% of Americans used AI daily for both work and personal life in late 2025.
- Nearly two-thirds call themselves early adopters, versus 30% of the public.
- They still support government involvement at 74%, close to the national 71%.
What it means
- On most questions, AI did not heavily divide Americans along partisan, geographic or educational lines.
- Americans want AI's benefits, fear the disruption, and want accountability from the companies building it.
- Anthropic plans to repeat the survey regularly and expand beyond the US.
Source
- https://www.anthropic.com/news/anthropic-public-record
- Published: 2026-06-12
- All figures are as stated in the source. Summary by this site.