AI Governance: What Americans Really Want
Research
•
Where the public sees the path forward, where consensus breaks, and the trade-offs they’re willing to make
An Editorial Note from Fathom
In 2024, Fathom’s first national survey on AI found an American public that was adopting the technology rapidly, yet feeling ambivalent about its trajectory and looking for a credible actor to establish guardrails. The headline finding was that Americans wanted guardrails—overwhelmingly, and across party lines—but had little confidence in government or industry to deliver them. Our second survey, conducted in fall 2025, found that trust gap holding steady—Americans were using AI more, but confidence in traditional institutions to govern it had not improved.
This report asks a harder question: Americans want governance, but what does their preferred AI future look like, and what trade-offs are they willing to make to secure it?
At this year’s Ashby Workshops, trust emerged as a central theme in every conversation about AI’s future—across labor, national security, education, and governance. This report puts empirical weight behind that call. It maps where trust is strong enough to build on, where it is fragile, and what it means for policymakers designing the governance frameworks that will need to be built in the years ahead.
This report’s findings are a call to action: Americans are not asking whether AI should be governed; they are asking who will govern it, whether it will reflect their priorities, and whether it will happen before the technology outpaces our institutions.
Executive Summary
Fathom’s mission is to build a governance architecture for the AI century—and that starts with understanding not just how people are using and thinking about this technology today, but what they’re willing to trade off to secure the AI future they want. Our latest survey, “AI Governance: What Americans Really Want,” goes beyond standard opinion polling to do exactly that.
Five Key Findings
1. There is a robust and durable mandate for trust and accountability.
Child safety, corporate accountability, and verifiable standards are Americans’ top priorities for a good future with AI. These priorities hold up across party lines, and even when the trade-offs are made explicit.
2. Consensus breaks where competitiveness is at risk.
Americans want safeguards on AI—but not at the cost of speed, control, or American leadership. Support for international AGI coordination drops sharply when ceding US control is made explicit. On child safety, the public is split between requiring verification before launch and allowing faster deployment with ongoing monitoring. The public wants governance and American leadership—and policymakers will have to design frameworks that reconcile the two.
3. Americans want workforce protection—and are open to a range of approaches.
From retraining programs to sovereign wealth funds that share AI-generated wealth with the public, every workforce policy tested in this survey commanded majority support. Americans decisively reject leaving workforce transition to market forces. There is broad demand but no preferred solution, a policy window that is open now but won’t be indefinitely.
4. Support for equitable access is wide but shallow.
Americans broadly agree AI’s benefits should be shared—but priorities like public AI infrastructure, subsidized access for schools and nonprofits, and community input in AI decisions rank well below trust and accountability. These are “yes in principle, not my priority” positions—mobilizing support around these principles will need a clearer urgency and benefit framing.
5. The public trusts third parties over government or industry.
Independent experts and nonprofits are the most trusted to ensure AI guardrails—ahead of tech companies, federal agencies, and elected officials. Americans see the barriers to progress on safeguards as largely political, not technical.