# Anthropic publishes its first Public Record survey results > The first wave surveyed nearly 52,000 Americans about hopes, fears, and regulation, creating a baseline whose methodology matters as much as totals. Canonical URL: https://www.devobs.io/news/news-first-anthropic-public-record-results/ By: Elias Brooks Published: 2026-09-06T11:58:54.626Z Updated: 2026-09-06T11:58:54.626Z Event date: 2026-06-12 Section: Data Anthropic published results from the first wave of its Anthropic Public Record survey series on June 12. The [announcement](https://www.anthropic.com/news/anthropic-public-record) says the wave was fielded in November and December 2025 and included nearly 52,000 Americans. It asked respondents about their hopes and fears for AI and their views on government involvement. ## The release establishes a recurring measurement surface Anthropic reports disease cures as the most common top-three hope and AI-related job loss as the most common fear in every state. It also reports broad support for some government role in AI regulation. These are survey findings, not forecasts of technological outcomes or evidence that respondents share one policy prescription. For data teams, the value of a large sample depends on recruitment, weighting, question wording, ordering, nonresponse, and the uncertainty of subgroup estimates. A national percentage and a state-level comparison can require different weighting and minimum sample rules. Readers should use the linked methodology before repeating a number, especially when converting ranked choices into claims about intensity. ## Archive the questionnaire with every chart A recurring survey becomes more useful when changes can be separated from instrument drift. Analysts should preserve wave dates, questionnaire version, response options, sampling frame, weights, exclusions, and code used to create published estimates. If future waves alter wording, maintain a comparability table rather than drawing a smooth trend line across incompatible measures. Engineering teams building dashboards from the series should expose sample size and uncertainty beside each filter, prevent unstable slices, and link every visualization to the relevant wave. Begin by reproducing one headline figure from the public methodology and data. If it cannot be reconstructed, treat the release as a reported result rather than a reusable dataset. ## Source references -