Bottom line: Broad American consensus: The public accepts AI benefits but demands government oversight and corporate liability, with only 15 % trusting AI companies themselves.
Anthropic has for the first time surveyed a representative sample of 52,000 Americans about their hopes and concerns around AI. The result: broad consensus across party and educational lines — with a clear demand for government oversight and developer accountability.
The first wave of the “Anthropic Public Record” was conducted between November and December 2025 with 51,993 Americans. 48 % cited curing diseases such as cancer or Alzheimer’s as one of their top three hopes for AI; 36 % hope for better support for people with disabilities. Technological advancement and general improvement of life both follow at 23 %.
On the concerns side, job loss dominates across all states: 64 % of respondents name this as their primary fear. Behind it ranks cognitive dependency (56 %) and disinformation (52 %). For CEOs, this is critical: only 15 % of Americans trust AI companies to make decisions about AI development on their own. 70 % of all respondents support government regulation (across party lines). When asked about effective protection, Americans see two levers as paramount: holding companies liable (47 %) and prioritizing safety over growth (44 %).
The broadest demands focus on data protection (56 %), child safety (52 %) and liability for damages (49 %). The remarkable finding: on most questions, there are no sharp divisions based on party affiliation, geographic location or educational level — there is broad agreement. The public wants AI benefits but fears disruption and demands accountability.
Anthropic plans to repeat the survey regularly and later expand it outside the US. The company combines this representative survey with ongoing user studies (81,000 Claude users) and data from the Anthropic Economic Index based on anonymized usage data.
Source: www.anthropic.com · Published June 11, 2026
Lumi AI News — AI-assisted curation pursuant to Article 50 EU AI Act. Paraphrase and classification via Lumi News Pipeline v1.7.1.