Britons doubt banks can spot sophisticated fraudsters

Digidentity has published research showing that 69% of Britons are not fully confident their bank could distinguish them from a sophisticated fraudster. The findings come as official figures show a 13% rise in bank and credit account fraud.

The survey of 1,000 UK adults found that 80% expect their bank to detect an AI-generated video or deepfake used to pass a selfie check. It also found that 53% believe a text or email one-time passcode alone is enough to prove someone's identity.

The results suggest a widening gap between what customers expect banks to prevent and what many understand about the limits of common security checks. They also indicate that banks are under pressure to strengthen identity verification without adding too much friction for account holders.

More than four in five respondents, or 81%, said banks should be held responsible if an identity verification provider fails to stop fraud. Another 78% said they would feel more confident using a bank with advanced identity verification.

At the same time, customers showed limited tolerance for longer checks. Some 62% said they would accept no more than two additional minutes of checks to reduce impersonation risk significantly, while 79% expect smoother identity checks within five years.

The data comes amid broader concern about AI's role in online scams and impersonation. Bank of England Governor Andrew Bailey has called on social media platforms to do more to tackle scams, and the Digidentity findings suggest consumers also expect banks to identify when the same technology is used against them.

Consumer expectations

Jonathan Evans, UK Country Manager at Digidentity, said: "If a criminal gains control of your email, the security code meant to protect you could land straight in their hands. Banks cannot assume that getting through a check means the right person has access. Deepfakes create a similar challenge for selfie checks, which must establish that the real customer is present rather than accept a convincing digital copy of their face.

"When a bank mistakes an impostor for its customer, the damage can follow that person long after the fraud is discovered. They could spend months or even years disputing debts and repeatedly proving they had nothing to do with accounts opened in their name."

This creates a practical problem for lenders and current account providers. Security tools that customers see as routine, such as one-time passcodes, may not be enough when criminals can take over email accounts or mobile devices. Face-based checks may also come under pressure from increasingly convincing synthetic media.

In customers' eyes, accountability still rests with the bank, even when external suppliers handle parts of the verification process. That could sharpen scrutiny of how banks select and oversee identity technology providers.

Bank responsibility

Evans said: "Banks put their own reputation behind the identity provider they choose. If that provider lets an impostor through, it is the bank's relationship with its customer that suffers.

"But stronger protection cannot mean an obstacle course for genuine customers. Making people spend more time proving themselves is not, on its own, evidence of better security. Banks should expect their provider to make impersonation harder while making life easier for legitimate customers."

The survey was conducted by Censuswide among a nationally representative sample of 1,000 UK adults aged 18 and over. The findings add to evidence that fraud prevention is becoming as much a matter of customer trust as of technology and operations, particularly as AI tools make impersonation attempts harder to detect.

For banks, that creates a difficult balance: meeting rising expectations for protection against deepfakes and account takeovers while keeping checks fast enough to avoid customers abandoning applications or routine transactions. The research suggests many customers will accept only a small increase in verification time, even as they expect banks to stop increasingly sophisticated attacks.