A poll conducted by the Center for Shared AI Prosperity (CSAIP) indicates that most Americans do not find voluntary industry agreements adequate for governing artificial intelligence (AI).
The survey sampled 2,498 respondents. According to CSAIP, 61% of those surveyed said that companies’ agreements to self-police AI are “not enough.” Among respondents who supported Donald Trump, 53% expressed the same view.
When asked whether the government should set and enforce rules for AI, 54% of voters agreed. The responses were split along partisan lines in the poll’s breakdown: 61% among respondents identified with Kamala Harris and 48% among those identified with Donald Trump.
Why this matters
The poll highlights a gap between public expectations and current policy approaches. The Trump administration and several leading AI firms, including Anthropic and OpenAI, have announced voluntary commitments intended to promote industry self-regulation. CSAIP’s findings suggest that a substantial portion of the public remains skeptical that voluntary measures alone will be sufficient, and many favor binding government rules.
Political implications
The results point to a potential mismatch between voter preferences for stricter AI oversight and the positions of elected officials in Washington. That asymmetry could generate political pressure for more comprehensive regulatory action, raising questions about how policymakers will respond if public demand for government-imposed rules intensifies.
The CSAIP poll results were published on the social media platform X.
