Summary

  • Almost 50% of Americans are optimistic that AI can help cure diseases; however, job loss remains the primary concern across various demographics.
  • Public trust in AI companies is alarmingly low, with only 15% expressing confidence in their decision-making regarding AI development, even less than trust in the federal government.
  • Support for AI regulation is widespread and crosses party lines, with over 70% of Americans advocating for government oversight, particularly concerning privacy, child safety, and corporate accountability.

A comprehensive poll conducted by Anthropic—the creator of Claude, which recently announced plans to go public—reveals that Americans have mixed feelings about artificial intelligence. While they are hopeful about AI’s potential to treat diseases, they are also concerned about the economic and social upheaval it could cause, and they harbor significant skepticism toward the companies developing these technologies.

The survey, titled "Public Record," collected insights from nearly 52,000 Americans in late 2025, making it one of the most extensive national surveys on public attitudes toward AI.

Concerns about job loss due to AI were prevalent, with 64% of participants expressing fears of displacement. This anxiety was consistent among both political parties and across all surveyed states, with heightened concerns among those with higher education levels, whose jobs are more likely to be affected by AI advancements.

On a more positive note, almost half of the respondents indicated that they wish for AI to help cure diseases such as cancer or Alzheimer’s, with 36% expressing a desire for AI to assist individuals with disabilities. Hopes that AI could replace human interaction, such as providing therapy or alleviating loneliness, ranked the lowest among the options given.

Despite these optimistic views, trust in the AI industry is notably low. Just 15% of those surveyed trust AI companies to make appropriate decisions about the technology's development and application, which is lower than trust levels in federal, state, and local governments, as well as international organizations; in contrast, independent experts garnered a 43% trust rate.

There is a strong bipartisan consensus on the need for government regulation, with over 70% of participants advocating for governmental involvement in AI oversight. Support for regulation ranges from 79% among Democrats to 68% among Republicans, with a strong emphasis on issues such as privacy, child safety, and corporate liability for any harm caused.

Anthropic has stated its intention to conduct this survey regularly and to broaden its scope beyond the United States.