European banks see fraud trust costs overtake direct losses, report says

European financial institutions now rank damage to customer trust above direct monetary losses as the main consequence of fraud, according to the European Fraud Insights Report 2026 cited by Stanley Epstein, associate at Citadel Advantage Group. The report says 42% of institutions identify customer attrition and trust erosion as fraud’s primary impact, compared with 36% that point to direct financial losses.

The findings frame fraud as a wider operating and reputational risk for banks, particularly as digital account switching reduces the friction for dissatisfied customers. Epstein wrote in a Finextra community opinion post that banks are facing more automated and personalised attacks, supported by fraud-as-a-service models and generative artificial intelligence.

The report also cites regional differences. In the Nordics, 60% of respondents said the effect of fraud on trust was their leading concern. Across Europe, 38% of banks reported higher operating costs linked to fraud, while 37% cited increased regulatory pressure.

Human checks remain central

Despite rising investment in automated detection, the report identifies human intervention by fraud specialists as the most effective control, giving it an effectiveness score of 1.94. Epstein said the reason is that some social engineering scams can lead victims to authorise payments in ways that appear normal to transaction-monitoring systems.

In that model, a scammer coaches the customer through a payment, so the transfer may match the customer’s device, behaviour and authentication credentials. A trained specialist can intervene through a call or in-person discussion and look for signs that the customer is being manipulated, according to the account cited by Epstein.

Santander UK’s “Break the Spell” team is cited as an example of that approach. According to the figures provided, the specialist team saved customers £6.8mn in 2025 and £24.4mn in total since 2021 by intervening in social engineering scams through phone and face-to-face contact.

AI changes attack patterns

The report says banks are preparing for more AI-initiated payment fraud, including systems that can imitate voices, faces and transaction histories. In the UK, 47% of banks expect significant growth in AI-initiated payment fraud over the next three years, 14 percentage points above the European average, according to the report.

UK financial services firms are also using AI defensively. The report says 75% have deployed AI for fraud prevention, while attackers are using generative tools to create counterfeit documents, deepfakes and more targeted scams.

Money mule accounts remain a key vulnerability. The report says the UK leads Europe in linking onboarding data with payment data in real time, at 33%, while the European average is 16%. That linkage matters because mule accounts may be opened with synthetic identities that look legitimate and then used to receive and move criminal proceeds before investigators can respond.

Banks move toward fewer platforms

The report says 81% of European banks are either consolidating fraud risk engines into a single platform or planning to do so. Of that group, 27% are already consolidating, while 54% plan to move multiple systems into one platform.

The barriers are mainly internal and technical, according to the report. It says 18% of banks face data quality or access limits, 17% cite organisational separation between fraud, technology and compliance teams, and 16% are slowed by legacy systems.

Dave Laramy, global head of fraud at Klarna and formerly of Danske Bank, said in the report that bringing data into one place can improve coverage across the customer lifecycle, while also making contingency planning and incident management more significant when a bank relies on a single platform.

This story draws on original reporting from Finextra Research.