European banks are no longer asking whether artificial intelligence belongs in financial crime and compliance. According to a new industry report, the debate has shifted to how quickly and safely AI can be scaled across fraud prevention, anti-money laundering and regulatory reporting, with clear implications for payments, digital money and future public infrastructure such as the digital euro.
The study, published by Chartis Research in collaboration with Hawk, finds that almost all banks now use AI in some form across their financial crime and compliance functions. However, most deployments remain limited to pilots or early production use, highlighting a gap between ambition and operational maturity.
Based on a survey of 125 financial institutions worldwide, including a significant European cohort, the report describes an industry in transition. Nearly half of respondents are still testing or piloting AI, while only a small minority use it at scale. Fraud prevention is the clear frontrunner, with AI already embedded in day-to-day operations at many banks, while AML transaction monitoring and regulatory reporting lag behind due to stricter explainability and governance requirements.
From experimentation to infrastructure
Traditional machine learning and natural language processing dominate current use cases, reflecting banks’ preference for proven, auditable techniques. More advanced tools such as generative AI and agentic AI are still rare in live compliance environments, but investment plans point to rapid change. Over 90 percent of surveyed banks plan to increase spending on generative AI in the next two to three years, and more than 80 percent expect to invest more in agentic systems.
That trajectory matters for Europe’s broader digital finance agenda. As instant payments, tokenised assets and a potential digital euro place new demands on transaction monitoring and fraud controls, banks are under pressure to modernise their compliance infrastructure without undermining trust or regulatory oversight.
The report suggests AI is already delivering measurable benefits. Almost half of respondents said AI saved their organisation more than one million dollars in the past year, largely through reduced false positives, faster investigations and lower manual workload. These efficiencies are particularly relevant as banks face rising compliance costs and increasingly real-time payment environments.
Regulation remains the friction point
Despite growing confidence, regulation continues to shape how far and how fast AI can be deployed. While three-quarters of respondents expect regulators to be broadly supportive of AI going forward, uncertainty over supervisory expectations remains a major constraint, especially in Europe’s rules-based regulatory environment.
Banks cite shortages of internal expertise as their biggest business challenge, closely followed by poor data quality and concerns around model governance. These issues become more acute with more autonomous AI systems, where auditability, accountability and human oversight are harder to guarantee.
This tension is likely to be familiar to European policymakers. The same questions arise in debates over the digital euro, where automation, privacy safeguards and control mechanisms must coexist. The report underlines that technological capability alone is not enough. Institutional readiness, clear governance frameworks and skilled teams are just as critical.
Why this matters for digital money
For the digital euro and other public digital payment initiatives, the findings offer a reality check. As transaction volumes accelerate and fraud techniques evolve, AI will be essential to safeguarding trust. But the banking sector’s cautious, incremental approach suggests that compliance-grade AI will advance more slowly than consumer-facing innovation.
The next phase, according to Chartis, will not be defined by whether banks adopt AI, but by how effectively they integrate it into core financial infrastructure under regulatory scrutiny. For Europe, that balance between innovation and control may prove decisive not just for banks, but for the credibility of digital money itself.