Presented by Veriff
Americans can’t reliably distinguish real content from AI-generated content, and that’s not just a media literacy problem; It is a direct threat to the way companies verify identity online.
New research finds that while many people are aware of deepfakes, their ability to distinguish them from reality is barely better than flipping a coin. A 2026 survey by Veriff and Kantar of 3,000 respondents in the United States, the United Kingdom, and Brazil shows that Americans score just 0.07 on a scale where 0 represents random guesses.
If people cannot distinguish authentic visual content, they will not be able to reliably distinguish authentic identities. In practice, that means that the same users who interact with digital services often cannot tell if the person on the other side of the screen is real.
That inefficiency has direct consequences for all digital businesses that rely on image- and video-based identity verification to confirm who is on the other side of a screen. This includes everything from customer bank onboarding and account recovery to marketplace seller verification, high-value e-commerce transactions, social platform authentication, and enterprise access control.
In the United States, those consequences are already material: synthetic identity fraud now accounts for billions in annual losses, and the tools to generate convincing fakes are now widely accessible.
The report also identifies a small but high-risk cohort: about 7% of users who perform poorly at detecting deepfakes, but remain confident in their ability and rarely verify what they see. While this is small as a percentage, at scale it represents millions of accounts that are highly exploitable targets for fraud.
If users cannot reliably distinguish real from synthetic identities, then any system that relies on visual verification is fundamentally exposed. Identity verification can no longer be considered a compliance function; instead, it must be built as a core digital infrastructure.
“Now that AI-generated content is becoming indistinguishable from reality, the human eye alone is no longer a reliable line of defense,” says Ira Bondar-Mucci, fraud platform lead at Veriff. "Companies and policymakers in the United States urgently need to close this awareness gap while investing in automated verification technologies that can detect what humans simply cannot."
U.S. Deepfake Awareness Gap Wider Than Expected
The United States may be the global epicenter of generative AI development, but American consumers demonstrate the least familiarity with deepfakes among the three markets surveyed. Only 63% of American adults are familiar with the term, compared to 74% in the United Kingdom and 67% in Brazil.
“There is a paradox at play,” says Bondar-Mucci. “The United States is the global epicenter of AI development, yet American consumers are the least familiar with one of its most dangerous byproducts. Historically, consumers have had greater trust in digital content, and the conversation around fraud focused more on data privacy than the authenticity of the content. The problem is that a lack of knowledge does not reduce the risk, but rather amplifies it. If you don’t know what a deepfake is, you are much less likely to stop and check if you you’ve found one.”
Human deepfake detection is barely better than flipping a coin
In practice, the randomness that characterizes the consumer’s ability to distinguish real from fake is evident in the way people evaluate different types of content. Video content proved to be especially difficult to evaluate, with fake videos often being flagged as authentic and real videos often being flagged as fake. Even in side-by-side comparisons, respondents split their judgments almost evenly, another indication that visual inspection alone is no longer a reliable method of verifying authenticity.
Overreliance on deepfake detection creates dangerous vulnerability
About half of American respondents say they are confident in their ability to identify deepfakes, but that confidence far outweighs actual performance, showing that self-assessment is meaningless in practice.
Within that population, there is that small but high-risk cohort: about 7% of users who are inaccurate, but overconfident in their ability and rarely verify suspicious content.
“This gap between trust and competence creates a false sense of security that scammers are willing to use,” says Bondar-Mucci. “When people believe they can’t be fooled, they stop looking for the signs. That’s precisely when they are most vulnerable, whether to a synthetic identity used in financial fraud or a fabricated video designed to manipulate trust.”
For businesses, the implication is clear: any organization that still relies on manual review or customer self-certification processes is inheriting this vulnerability directly. Human judgment is an increasingly unreliable safeguard and verification must be built into systems by default. This means automated, technology-based, and not dependent on the end user’s self-assessment of their ability to distinguish real from fake.
Americans are worried about deepfakes, but trust platforms to handle them
Concern about deepfakes is high across the United States, with 79% of respondents saying they are somewhat or extremely concerned about personal fraud and phishing.
The United States differs from other markets in where that concern is headed. Americans are more likely than respondents in the United Kingdom or Brazil to rely on social media platforms and digital services to identify and manage AI-generated content. That delegation of responsibility may be reducing individual vigilance at exactly the moment when the threat is accelerating.
“We are seeing the use of synthetic identities to open fraudulent accounts and authorize transactions, and deepfake videos deployed to bypass basic verification checks,” he explains. “What makes this particularly urgent is the combination of high concern with relatively high trust in the platform. That gap between perceived and actual protection is exactly where fraud thrives.”
The business case for automated identity verification has never been stronger
The gap between what Americans think they can detect and what they can actually detect is not a knowledge problem that awareness campaigns will solve, but a design flaw in any system that places the burden of identity verification on unaided human judgment.
The effective response is not to remove humans from the verification loop, but to stop assigning them tasks that human perception can no longer reliably perform. Organizations that persist in relying on manual review or customer self-certification processes are absorbing this vulnerability into their operations.
The alternative is AI-powered automated identity verification that operates at the point of interaction, detects synthetic media before a human decision is required, and does not rely on the end-user’s ability to distinguish real from fake.
“Seeing is no longer believing,” says Bondar-Mucci. “Companies that build verification infrastructure around that reality, rather than assuming it will be otherwise, are the best positioned to maintain customer trust as the synthetic media landscape continues to evolve.”
Sponsored articles are content produced by a company that pays to publish or has a business relationship with VentureBeat, and are always clearly marked. For more information, contact sales@venturebeat.com.
