Notifications
Clear all
Topic starter
11/07/2026 8:30 pm
TL;DR: AI is lowering the cost of check fraud by making fabricated paper checks look routine enough to pass human review, while about two-thirds of businesses encountered check fraud in 2024, according to the 2025 AFP Payments Fraud and Control Survey Report. The control problem is no longer just anomaly detection, but deciding what to trust when fraud is designed to resemble normal operations.
NHIMG editorial — based on content published by Secureframe: What a Fake Check Taught Us About AI-Enabled Banking Fraud
By the numbers:
- About two-thirds of businesses encountered check fraud in 2024, making it the most common fraud method that year.
- When AWS credentials are exposed publicly, attackers attempt access within an average of 17 minutes.
Questions worth separating out
Q: What breaks when AI-generated fraud looks routine enough to pass review?
A: Review processes break when they rely on human pattern recognition to separate legitimate from malicious items.
Q: Why do normal-looking payment artefacts create such a governance problem?
A: Normal-looking artefacts are dangerous because many controls assume suspicious activity will stand out.
Q: How do banks know if their fraud controls are actually working?
A: They should test whether suspicious transactions are declined or challenged in real time, whether payee verification stops redirection attempts, and whether risky sessions are suspended when the runtime environment changes.
Practitioner guidance
- Harden exception approval paths Require a second control owner for high-risk payment exceptions and separate initial review from final approval.
- Shift from appearance checks to provenance checks Validate who issued the payment, which account authorised it, and whether the transaction aligns with expected business behaviour.
- Reclassify paper checks as a higher-risk channel Update fraud risk registers so low-volume paper checks receive explicit monitoring, escalation, and recovery playbooks rather than default safeguards only.
What's in the full article
Secureframe's full blog covers the operational detail this post intentionally leaves for the source:
- How the fabricated checks were constructed and why they passed routine scrutiny.
- The specific control assumptions behind Positive Pay and similar exception workflows.
- Why visual plausibility creates a blind spot in payment review processes.
- Practical examples of how businesses can adapt escalation paths when fraud looks ordinary.
👉 Read Secureframe's analysis of how AI is changing banking fraud →
AI-enabled banking fraud: what banks and finance teams must change?
Explore further
11/07/2026 8:59 pm
Ordinary fraud has become the dominant control problem. The article shows that AI-enabled fraud does not need to look dramatic, only routine enough to pass the threshold for human acceptance. That shifts the security question from whether fraud is sophisticated to whether the control environment is still built around visual suspicion. For banks and finance teams, the practitioner conclusion is that normality must now be treated as a threat condition, not a comfort signal.
A question worth separating out:
Q: Who is accountable when AI-driven fraud bypasses identity controls?
A: Accountability usually sits across IAM, fraud operations, and product security, because the failure spans authentication, session trust, and abuse response. If the organisation cannot explain why an automated actor was treated as trustworthy, the gap is governance, not just detection. That is the level leaders should review.
👉 Read our full editorial: AI-enabled banking fraud is making ordinary checks unsafe
