Who Regulates AI? Americans Don’t Trust Their Own Answer

One-line job: Map who actually governs AI today — and why American confidence in that governance is the lowest measured anywhere. Audience: Operators trying to figure out which rules will bind them, and policy watchers tracking the trust layer. Not for: Readers wanting legal advice. This is the public-opinion and regulatory-landscape picture. Last verified: 2026-08-13 Evidence weight: documentation-verified

Here’s the number that should reframe every “regulation will slow us down” conversation in America: only 31% of Americans trust their own government to regulate AI appropriately — the lowest figure of any country surveyed, against a 54% global average (Stanford HAI’s 2026 AI Index, citing Pew 25-country data).

And the direction of the worry matters as much as its size: 41% of Americans say federal AI regulation will not go far enough, versus 27% who fear it will go too far. The American public isn’t bracing for overreach. It’s waiting for someone to show up.

The cross-country picture

Pew’s October 2025 survey across 25 countries found sharply asymmetric regulator trust: a median of 53% trust the EU to regulate AI well, 37% the U.S., and 27% China. Within the U.S., partisan splits run deep too — Republicans 54% vs. Democrats 36% on trusting their own country’s approach.

So when American companies complain about regulatory uncertainty, the polling says their own public largely agrees the current framework is inadequate — not excessive.

What actually exists today

The binding statute with real teeth is European. The EU AI Act (Regulation 2024/1689) is in force with:

• Prohibited practices (Article 5): including untargeted facial scraping, emotion inference in workplaces/schools
• High-risk obligations across employment, education, and biometric categories
• Penalties up to €35M or 7% of worldwide annual turnover

In the United States, governance remains patchwork: sector-specific enforcement, state-level experiments, executive actions of varying durability. Nothing resembling the EU’s comprehensive statute exists federally.

Why this matters commercially

Three practical consequences fall out of the trust data:

1. The EU framework de facto applies globally to anyone with European users — and it’s the strictest regime, so it sets the engineering floor.
2. American companies can’t count on domestic regulatory legitimacy as a shield. With 31% trusting the regulator, “the government allows it” persuades almost nobody.
3. The 41%-want-more-regulation number is a clock. When the public wants more rules than it has, rules are coming; the uncertainty window closes from one side.

The call

1. Build to the EU AI Act floor regardless of where you’re headquartered. It’s the only comprehensive statute, it reaches extraterritorially, and it’s what “careful” looks like in enforceable terms.
2. Track state-level U.S. requirements separately — patchwork means your compliance map has fifty tiles.
3. In your user communications, don’t cite regulatory approval as reassurance. The polling says most Americans don’t find the regulator credible in the first place.
4. Budget for compliance as a product feature, not legal overhead. Under the EU penalties schedule, it’s priced accordingly.