Suspicious Gambling Transactions Up 4.5x as AI Fraud Bites

What the data shows

Sumsub's third annual iGaming Fraud Report puts a hard number on something operators have been feeling in their fraud queues: suspicious transaction volumes grew 4.5 times over the twelve months to Q1 2026. That's not incremental drift — it's a step-change in the scale of what compliance and payments teams are being asked to absorb.

AI is doing the heavy lifting for fraudsters

At the Payment Expert Summit at SBC Americas, GeoComply CEO Kip Levin told attendees that organised groups are now using AI to accelerate attacks on regulated betting platforms, with particular pressure emerging around deposit flows. The implication is straightforward: fraud that once required significant human coordination can now be industrialised. Volumes go up, attack patterns shift faster, and the window between a new fraud vector appearing and it being mitigated gets shorter.

Put the two data points together and a clear picture forms — it's not just that fraud is more common, it's that the tooling available to bad actors has made scaling attacks dramatically cheaper.

Why operators can't treat this as a back-office problem

Payment fraud in gambling isn't only a financial loss issue. Chargebacks damage processor relationships, suspicious transaction flags can trigger regulatory scrutiny, and AML failures carry licence consequences. When volumes quadruple in a year, any compliance stack that was sized for the old normal is already behind.

Key pressure points operators should be stress-testing right now:

• Deposit-side monitoring: organised AI-driven attacks are targeting deposit flows specifically, per the SBC Americas commentary
• Transaction velocity rules: static rule sets struggle when attack patterns are being iterated by automated tooling
• AML alert triage capacity: a 4.5x volume increase hits human review teams hard unless automation is absorbing the first-pass work
• Vendor SLAs: whether fraud tooling is in-house or outsourced, response-time commitments need revisiting against current threat volumes

The operator takeaway

Boards and CFOs tend to treat fraud infrastructure as a cost centre to be managed down. The Sumsub figures, set alongside the AI-acceleration point from GeoComply's CEO, make the counter-argument: under-investing now means catching up under regulatory or financial duress later. Operators who haven't reviewed their payment fraud and AML stack in the last twelve months are almost certainly working with assumptions that no longer reflect the threat environment.