Most Americans can’t tell a deepfake from reality – even when they’re sure they can

New York – May 20, 2026 – A new report from Veriff, the global AI-native identity platform, exposes a troubling gap between Americans’ confidence in spotting deepfakes and their actual ability to do so. Veriff Deepfakes Report 2026, produced in partnership with Kantar, surveyed 1,000 U.S. adults as part of a broader 3,000-person study spanning the United States, United Kingdom, and Brazil.

The findings paint a clear picture: deepfake detection, for the average individual, is effectively a coin flip.

Detection is barely better than guessing

When participants were shown a series of 16 visuals – eight real, eight AI-generated or manipulated – U.S. respondents averaged a detection score of just 0.07 on a scale of -1 to 1, where 0 represents pure chance. That score is statistically indistinguishable from random guessing.

The breakdown tells the same story:

• 14% scored in the lowest possible range
• 16% performed worse than chance
• 15% landed exactly at chance level
• 38% performed slightly above chance
• 18% achieved the highest scores

Video content proved particularly deceptive. In one side-by-side comparison of a male and female video pair, respondents were split nearly evenly on the male pair (52% correct), but got the female pair badly wrong: 70% misidentified the deepfake as the real video.

AI-generated images of women and faceswap visuals were also highly deceptive, while results varied more for male subjects.

Americans show the lowest level of awareness

Despite being home to the world’s leading AI companies, the U.S. had the lowest level of deepfake awareness among the three markets surveyed. Only 63% of American adults are familiar with the term “deepfake”, compared to 74% in the UK and 67% in Brazil. And unlike most countries, younger Americans are no more aware of deepfakes than older generations.

At the same time, the U.S. stands apart in a different way: Americans are more likely than their UK and Brazilian counterparts to trust social media platforms and digital services to handle AI-generated content on their behalf. That combination, lower personal awareness alongside higher platform reliance, creates a vulnerability that fraudsters are well-positioned to exploit.

Half of Americans think they can spot a fake, but they’re wrong

Roughly half of U.S. respondents described themselves as confident in their ability to identify manipulated media. But confidence and accuracy are not the same thing. While slightly more confident respondents did perform better on average, the gap was small, and even the most self-assured participants scored far below what would constitute reliable detection.

The most common tactics Americans reported using to spot deepfakes – unnatural skin (53%), oddities in appearance (52%), and unnatural movement or expressions in videos (51%) – are precisely the artifacts that modern AI tools are engineered to eliminate.

“Our research reveals what may be the most dangerous dynamic in the deepfake era: overconfidence,” said Ira Bondar-Mucci, Fraud Platform Lead at Veriff. “Seeing is no longer believing. The most dangerous element of this report isn’t that deepfakes are becoming increasingly sophisticated, but that people think they can tell, and they cannot.”

The high-risk segment

Across all three markets, roughly 7% of respondents fall into a “high-risk” category: people who perform poorly at detection, are highly confident they would catch a fake, and rarely or never verify suspicious content they encounter online. This group represents a persistent soft target for deepfake-driven fraud.

In the U.S., this pattern holds even among older adults — a notable contrast with the UK and Brazil, where older respondents were somewhat less likely to fall into this category. University-educated respondents were less likely to be high-risk across all markets.

Concern is high, but action is lagging

Despite the detection challenges, Americans are acutely aware of the stakes. 79% of U.S. respondents are rather or extremely concerned about deepfake-driven personal fraud and impersonation, making it the top fear in the survey. Political misinformation (77%) and the erosion of online trust (75%) follow closely behind.

The problem is that concern alone doesn’t translate into protection, particularly when nearly half the population believes platforms will handle the problem for them.

“That gap between perceived and actual protection is exactly where fraud thrives. For companies, the answer isn’t to reassure customers – it’s to earn that trust through action,” said Bondar-Mucci. “It means deploying AI-driven biometric authentication that can verify a real person in real time, detect synthetic media at the point of interaction, and do so without relying on the customer to spot the fake themselves. The deepfake arms race is an AI problem that requires an AI solution. The companies that build this partnership between human oversight and automated verification today will be the ones that earn and keep their customers’ trust tomorrow.”

Can you tell real from fake?

Take Veriff’s Deepfakes Quiz to test your own detection accuracy.

Methodology

The survey was conducted by Kantar in February 2026 using an online access panel. The study included 3,000 respondents aged 18–64, 1,000 each in the United States, United Kingdom, and Brazil, with nationally representative quotas applied for age, gender, and region. Participants assessed 16 visuals (8 real, 8 AI-generated or manipulated), including fully AI-generated images, AI-generated videos, and faceswap content. All visuals were shown in randomized order. Detection accuracy was calculated using a scoring index benchmarked against a defined chance-level baseline.

About Veriff

Veriff is a global AI-native identity platform that helps organizations build trust online. Leading companies across financial services, marketplaces, mobility, gig economy, and other digital sectors rely on Veriff’s technology to stay compliant, prevent fraud, protect users, and scale globally.

Veriff’s trust infrastructure supports the full customer journey, from verification to ongoing authentication and fraud prevention, with the least friction for honest people. Built for global scale, Veriff helps businesses expand across borders without the complexity of managing identity verification, compliance, and fraud in multiple markets – creating a single source of truth for trusted identities.