Introduction
In July 2025, the UK's Financial Conduct Authority fined Monzo £21.1 million. Among the findings: customers had opened accounts using blatantly false addresses including 10 Downing Street and Buckingham Palace. The checks ran. The records were filed. The answers were absurd on their face.
That is the uncomfortable truth about Know Your Customer. KYC verifies documents, not people and it verifies them once, on the day someone signs up, then never looks again. Meanwhile the attack has moved. US consumers lost $27.2 billion to identity fraud in 2024, a 19% rise in a single year. In the first half of 2025, 8.3% of digital account creations were suspected fraudulent roughly one in twelve. Global losses to synthetic identity fraud now run between $20 billion and $40 billion annually.
None of those attacks fail a document check. A rented passport is a real passport. A deepfaked selfie clears liveness. A mule account's ID belongs to a genuine person — one who was paid to sit through onboarding and hand over the login afterwards.
What fraudsters cannot fabricate is history. Every real identity leaves a trail across the internet. Every manufactured one arrives without one.
Conclusion
KYC is not the problem, and nothing here argues for dropping it. Where AML, CDD or right-to-work rules apply, they apply in full. The problem is asking a one-time document check to answer a question it was never built for: is the person behind this screen who they claim to be, today?
Digital footprinting answers that from data you already collect. Email age, phone line type, platform presence, device print, network origin, link velocity — six independent signals, none of them supplied by the applicant, all resolving in milliseconds before anyone uploads anything.
Run passively first, it changes the economics of onboarding. Low-risk users pass straight through. Medium-risk users meet a step-up. High-risk users are held before the payout, not after. Same regulation, unequal friction and the friction lands where the risk actually is.
Atna.AI builds precisely this layer. Our digital footprinting engine analyses device, IP and behaviour patterns in real time to catch early fraud signals with zero friction — no biometrics, no video calls, no extra screens. It feeds directly into predictive risk scoring and customer onboarding, so identity stops being a checkbox at signup and becomes a claim you keep testing.
