The new framework should improve firms’ risk management, but it also risks driving up their operational workloads, skill requirements, and demands on the second and third lines. Given that KYC is typically the costliest area of financial crime prevention, accounting for up to 5% of banks’ total expenses, obliged entities (OEs) face a clear risk of cost buildup and profitability pressure. That makes new tech-led approaches to KYC a necessity. The tools now available to OEs can not only help to deliver compliance, but also to catalyse wider structural improvements in risk and efficiency.
Many firms already apply workflow tools and robotic process automation to CDD, but fragmented silos, duplication, and manual workarounds mean that KYC activities often remain highly resource intensive.
Now though, rapid advances in AI and advanced analytics mean that firms can completely redesign KYC processes, reassigning activities from humans to machines. Above all, the use of agentic AI — not a single tool, but an ecosystem of specialised agents with distinct responsibilities — can reduce headcount requirements and allow AML experts to focus on value-adding activities.
Financial firms’ onboarding and CDD processes typically feature three key stages:
- Collection of customer data by the client facing channel
- Assessment and analysis, leading to an accept or reject decision
- Final validation by Compliance, or by dedicated AML centres for high-risk cases.
The greatest opportunities for tech-powered KYC enhancement lie in the core second stage. This is the most complicated and time consuming phase, with activities that include verifying data completeness, requesting additional documentation, searching information sources, and writing up customer dossiers. At present however, only 20-25% of staff time is typically spent on risk analysis or expert judgement. There is significant scope for analysis and assessment tasks to be conducted by AI agents, subject to review and oversight by human AML experts. That could include:
- Retrieving data from internal systems and third-party providers
- Identifying and requesting further documentation or evidence
- Screening names against sanction, PEP and crime lists
- Spotting hidden patterns and raising red flags
- Drafting customer intelligence dossiers
- Triggering escalation for unusual or high-risk cases
The EU’s new AML/CFT framework was drafted with the potential for greater non-human KYC activity in mind. It permits automated decision-making, including by AI, but with a clear requirement for real human involvement in every decision. The incoming regime provides a real opportunity for European institutions to leverage technology — improving KYC, reducing cost, and making better use of scarce human talent — to create a new ‘augmented compliance’ model that’s fit for the future.
Discover how the EU’s new AML framework will revolutionise KYC processes, harnessing AI and advanced analytics to tackle risk, drive efficiency, and transform compliance. Find out what these changes mean for financial institutions and their future success
