AI reshapes fraud prevention strategies

By Milan Rojan

Fraud prevention across the UK financial services sector is advancing unevenly, with differences in AI adoption and legacy infrastructure leaving some institutions more exposed to increasingly sophisticated financial crime, according to insights shared ahead of the NextGen FinCrime event.

The commentary draws on a forthcoming survey of 187 UK financial crime leaders, which found that many organisations have deployed or are planning AI-driven fraud detection, stronger authentication and data-led monitoring. However, the findings also suggest that adoption varies significantly between larger institutions and smaller or regional banks.

While major banks have increasingly explored AI across fraud detection, customer communications and decision support, many smaller institutions have continued to rely on legacy systems, slowing the adoption of advanced fraud prevention capabilities.

The survey has highlighted a broad threat landscape rather than a single dominant risk. Respondents identified email fraud, identity theft, payment fraud, Authorised Push Payment (APP) fraud and deepfake-enabled attacks among their primary concerns, reflecting the growing use of AI by cybercriminals.

The findings have aligned with figures from UK Finance, which reported that total payment fraud losses reached $1.71 billion in 2025, rising 4% year on year. APP fraud losses increased 19% to $772 million, accounting for almost half of all payment fraud losses, while investment fraud rose 40% to a record $297 million. Much of the increase was linked to scams exploiting consumer trust through online and telecommunications channels.

The analysis has suggested that while AI is strengthening fraud detection, financial institutions must also address governance, transparency and workforce readiness. Survey respondents identified phishing, credential theft, business email compromise and insider threats as leading attack vectors, yet employee training ranked among the lowest priorities for investment.

The article has argued that effective fraud prevention requires a combination of advanced technology, human oversight and organisational capability. As financial institutions accelerate AI adoption, balancing automation with governance and skilled personnel will be critical to improving resilience. The findings have suggested that narrowing the gap between technology leaders and slower adopters may prove essential as fraud techniques continue to evolve across the financial services sector.