Americans trust AI more than Congress

If you had to pick who you trust more to benefit society between AI, social media platforms, and Congress, who would you choose? According to a new study from Rutgers University, AI actually wins that contest.

Rutgers researchers polled Americans to understand how everyday people perceive artificial intelligence and AI-generated content in their daily lives. The overarching takeaway is that Americans have distinctly mixed feelings, though the breakdown varies wildly across demographics.

Trust beats Capitol Hill, but humans win the newsroom

About 47% of Americans trust AI to benefit the public. While that falls short of a majority, it easily beats out social media at 39% and Congress at 42%.

Demographics reveal sharp divides. Men trust AI at 52% compared to 43% of women. Younger adults aged 25 to 44 show the highest optimism at 55%. Geography matters too: 53% of urban dwellers trust AI, while only 38% of rural residents share that sentiment.

When it comes to corporate accountability, exactly 50% trust businesses to use AI responsibly. That confidence peaks at 65% among respondents with graduate degrees earning over $100,000 annually. Urbanites again proved more trusting at 53%, compared to 42% among rural populations.

However, humans retain their edge in information delivery. While 48% trust AI-generated news, 62% trust mainstream journalists, showing people still prefer human oversight when facts are on the line.

The confidence trap: a 3.3 out of 8 score

The study also exposed a significant gap between how much people think they know about AI versus reality.

Only 26% of Americans say they are very familiar with AI, while 63% admit they are not. Younger adults, men, and educated respondents reported the highest familiarity.

Younger crowds also felt more confident spotting synthetic content, but overall, just 13% of respondents felt very confident identifying AI-generated material. Another 30% felt somewhat confident.

When Rutgers gave participants an actual eight-question knowledge quiz about AI, the national average was just 3.3 out of 8. Even though higher education correlated with higher scores, general technical literacy remains surprisingly low across the board.

Why it matters

We think this poll highlights a fascinating reality: AI's relatively high trust numbers are less about flawless technology and more about how much people dislike legacy institutions. When an algorithm outscores elected lawmakers and social media feeds, the bar is admittedly on the floor.

Yet the dangerous takeaway is the mismatch between trust and literacy. When average Americans score 3.3 out of 8 on basic AI knowledge and only 13% feel truly confident identifying synthetic media, you have the exact recipe for deepfake scams and hallucination spread.

For enterprise tech builders, there is clear runway. High-earning, educated users are overwhelmingly ready to trust companies deploying AI responsibly. That validates premium software adoption in professional settings.

For publishers and media outlets, human reporting remains an enduring advantage. Mainstream journalists holding a 14-point trust lead over AI copy shows that deep reporting and verified provenance are still strong moats that synthetic content cannot easily displace.