Why AML and KYC Compliance Costs Keep Rising for Banks

Anti-money-laundering and know-your-customer compliance costs have climbed almost every year for over a decade, across every major banking market. That pattern isn't the result of any single regulation or enforcement wave. It comes from three features of how the work itself is structured, each of which resists automation for reasons that have nothing to do with how much a bank is willing to spend.

Why Transaction Monitoring Still Needs Human Review

Automated transaction-monitoring systems flag activity that deviates from a customer's typical pattern, but flagging isn't the same as deciding. A flagged transaction still has to be reviewed by a trained analyst who checks it against the customer's history, business type, and geography before deciding whether to file a suspicious activity report. Monitoring systems are tuned conservatively because missing a real case carries far heavier regulatory consequences than over-flagging, which means the volume of alerts requiring human judgment stays high even as the underlying software improves.

This is the main reason AI adoption in AML and KYC work has grown quickly, according to Fenergo's Financial Crime Industry Trends 2025 report, without compliance spending falling. Reported use of advanced AI tools rose from 42% of firms in 2024 to 82% in 2025. Those tools are helping analysts triage and prioritize alerts faster, but a human still has to make the final call on anything genuinely ambiguous, so the analyst headcount underneath the software has not shrunk to match.

Why Know-Your-Customer Never Actually Finishes

KYC is often described as an onboarding step, but for any existing customer it's a recurring obligation. Institutions are required to periodically refresh customer risk profiles, and specific events, a large unexplained deposit, a change in business activity, a customer's move to a higher-risk jurisdiction, can trigger a full re-verification outside that schedule. A bank with millions of retail and commercial accounts is running this refresh cycle continuously, not once per relationship.

That ongoing workload is a large part of why average annual AML/KYC spend varies by market even among developed economies with similar regulatory frameworks. Fenergo's 2025 survey, a single-year snapshot rather than a fixed benchmark, found UK institutions averaging $78.4 million a year, ahead of US firms at $72.2 million and Singapore firms at $68.2 million.

Why Operating Across Borders Multiplies the Work

A bank operating in multiple jurisdictions doesn't get to run one compliance program and apply it everywhere. Each country sets its own customer due-diligence thresholds, its own list of reportable transaction types, and its own definitions of high-risk customers and sanctioned entities. A multinational institution has to maintain separate compliance logic, and often separate staff, for each jurisdiction a customer touches, even when the customer is the same legal entity end to end. This is a structural cost that scales with a bank's geographic footprint rather than with any particular year's regulatory climate, which is why larger cross-border institutions consistently report higher per-firm compliance spend than domestic-only firms of similar size.

What a Compliance Program Still Misses

None of this guarantees that spending catches everything it's meant to. TD Bank had an active AML and KYC program in 2024 when regulators found it had approved more than $470 million in transactions tied to a Chinese money-laundering network, according to Thomson Reuters' overview of the case. The bank was fined more than $3 billion. A program can run every required check, staff every required review, and refresh every required customer profile, and still miss a pattern that only becomes visible once investigators look at it from outside the bank's own systems.

That's the tension sitting underneath the rising cost figures: institutions keep adding staff, tools, and jurisdiction-specific processes to a compliance function that, by its own design, evaluates transactions one flagged case at a time. Coordinated laundering networks are built to look unremarkable at that scale. Closing that gap would mean rethinking what the monitoring is looking for, not just how much money is spent looking.