National survey finds strong, bipartisan support for AI guardrails.
Americans want governance of AI systems, with strong support for safeguards centered on trust, safety, and accountability, according to the AI Governance: What Americans Really Want survey released by Fathom.
The report finds that the American public places a high priority on AI governance over other considerations, including innovation. When asked to rank what matters most for a “good future with AI,” respondents ranked these four top priorities as “very important.”
- Protecting children: 70% say AI products intended for children should be verified as safe before deployment
- Accountability: 67% say there should be clear paths for accountability when AI causes harm
- Safety standards: 62% say the public should be able to verify that AI systems meet clear safety standards
- Human-centered AI: 60% say AI should support human reasoning rather than replace it
Even when hard choices are introduced, these priorities remain stable. In forced-choice questions, respondents continue to favor governance measures.
- 77% favor verification of AI systems over avoiding bureaucratic delays
- 72% favor accountability even if it discourages innovation
Further, child safety, accountability, and trust rank as the top three priorities regardless of gender or political affiliation, with differences typically within a few percentage points.
US Competitiveness Softens The Stance
However, support for governance weakens when it conflicts with U.S. competitiveness. When governance measures are framed as limiting American control or leadership, public opinion becomes more divided. For example, in questions about international coordination on advanced AI systems, 49% support coordinating with other countries, compared with 41% who prioritize U.S. control.
This divide is more pronounced across political groups. Among Democrats, 59% support international coordination compared to 31% who favor U.S. control. Among Republicans, 51% favor U.S. control compared to 39% who support international coordination.
Americans Want Workers Protected
The study shows strong support for workforce protections tied to AI adoption. Respondents favor policies that help workers adapt rather than relying on market forces. The top three supported measures are:
- 84% support retraining and reskilling programs
- 83% want transferable benefits to help workers switch jobs
- 78% are in favor of expanding protections such as unemployment insurance and tax credits
Further, 66% of Americans want policies to protect workers, not the market. This holds true across party lines.
Who Enforces Governance?
Respondents put their trust in independent organizations:
- Independent experts: 71% trust (77% Democrats, 69% Republicans, 64% Independents)
- Technology companies: 61% trust (60% Democrats, 66% Republicans, 52% Independents)
- Federal government agencies: 51% trust (47% Democrats, 58% Republicans, 45% Independents)
Trend data shows trust in independent organizations has increased, rising from 54% in 2025 to 66% in 2026. Trust in technology companies also increased from 50% to 61% over the same period, while trust in government remains relatively unchanged.
What’s Stopping Us From Achieving Effective AI Governance?
The report identifies the perceived barriers to achieving AI governance priorities as political rather than technical.
- Conflicting interests among companies, government, and the public: 42%
- Lack of public understanding or engagement: 37%
- Cost or funding constraints: 23%
The survey is based on a nationally representative online sample of 2,036 respondents. It uses a multi-stage design, including ranking exercises, trade-off scenarios, and forced-choice comparisons, to measure how public preferences change under different conditions.
