Banks overrate fraud readiness as data gaps persist

Feedzai has published research showing a gap between banks' confidence in their fraud readiness and the systems they have in place. Institutions rated themselves at 80.2 out of 100, but scored 66.9 when assessed against operational factors.

The findings are based on a survey of 1,030 senior fraud and financial crime leaders at banks, fintechs and payment providers across 12 countries, alongside transaction data covering more than 120 billion transactions.

Despite those operational shortfalls, spending plans remain strong. Some 96% of respondents said they expected their budget for AI-based fraud tools to grow over the next two years.

Yet higher spending has not resolved basic issues around data access and workflow. Only 41% of respondents said they had fully connected data across fraud, anti-money laundering, compliance and cybersecurity teams, while 73% said their fraud teams spent too much time manually pulling data from different systems instead of analysing threats.

That disconnect affects day-to-day investigations. Analysts investigating a suspicious payment may need to check several separate systems before building a full customer picture, slowing intervention.

Readiness gap

The gap between self-assessed readiness and measured readiness was 13.3 points. The wider industry score in the index was 69.3 out of 100.

Only 25.5% of institutions were classed as "leading", requiring a readiness score of 76 or higher. A further 73% fell into the "Capable" tier, with scores between 51 and 75.

Institutions that ranked highest tended to have connected their data more effectively and reduced the amount of analyst time spent on manual information gathering.

The index assessed institutions across five equally weighted areas: leadership vision, network intelligence, data infrastructure, AI maturity and operational structure. The measured score excluded leadership vision because that element relied on self-assessment.

Sharing data

One notable finding was the level of support for cooperation between institutions. The survey found that 87% of respondents would consider securely sharing fraud intelligence with other institutions, including direct competitors.

The report linked that shift to the rise of scams in which customers are persuaded to send money themselves. It said scams are now the biggest source of fraud losses in 10 of the 12 countries surveyed.

Nuno Sebastião, Co-Founder, Chairman, and Chief Executive Officer at Feedzai, said: "Banks aren't short on ambition or budget. What's holding them back is inefficiency: data that doesn't connect and analysts spending their day hunting for information instead of stopping fraud with modern technology such as AI.

"Criminals don't target one bank at a time, so the industry can't defend itself one bank at a time either. The fact that most institutions would consider sharing intelligence with their competitors shows how much has changed in the fight against fraud."

AI adoption

The survey also pointed to growing interest in newer AI tools. Some 68% of institutions said they were using or planning agentic AI for three or more fraud use cases.

At the same time, respondents identified trust in autonomous decision-making as their main concern. That suggests institutions are prepared to invest in broader AI deployment but remain cautious about handing over decisions without human oversight.

The survey covered Australia, Brazil, Canada, Colombia, France, Germany, Mexico, Singapore, Spain, the United Arab Emirates, the UK and the US. Respondents included senior leaders responsible for fraud and financial crime functions across banks, fintechs and payment providers.

The findings suggest that, for many institutions, the main obstacle is no longer willingness to spend on fraud prevention tools but the difficulty of joining up internal systems and turning investment into faster operational responses. Only 41% have fully connected data across fraud, anti-money laundering, compliance and cybersecurity teams.