The global fraud prevention landscape is shifting in paradoxical directions: even as industry-wide identity fraud rates show modest improvement, the character of attacks has grown dramatically more sophisticated. Sumsub, the identity verification and compliance platform, has responded by launching a dedicated fraud prevention starter kit designed to give businesses a structured, data-driven framework for evaluating applicants at the moment of onboarding — precisely where the battle against fraud is most often won or lost.
The headline number driving this product launch is stark: complex attacks have risen by 180%, a figure that underscores a qualitative shift in the threat environment. Fraudsters are no longer relying on crude, easily detectable methods. Instead, they are deploying layered, multi-vector schemes that combine synthetic identity construction, document forgery, and increasingly convincing behavioral mimicry. The volume of unsophisticated attacks may be easier to absorb, but a 180% surge in complexity means that even well-resourced compliance teams face mounting pressure to triage applicants more effectively at scale.
Sumsub's starter kit addresses this challenge directly by leveraging onboarding data — the information businesses already collect during the customer acquisition process — to assign each applicant a risk classification of low, medium, or high. From that classification, compliance officers and operations teams can route each case through one of four defined pathways: outright approval, escalation for further checks, manual review, or outright blocking. The elegance of this architecture lies in its proportionality. Not every flagged applicant represents a genuine threat, and not every clean profile is free of risk. A tiered system allows institutions to concentrate their investigative resources on the cases that genuinely warrant them, rather than applying blanket scrutiny that slows legitimate customer acquisition.
The timing of this product launch is notable, though the underlying data sends a more nuanced signal. Sumsub's own internal data shows that the global identity fraud rate declined from 2.6% to 2.2% over the measured period — a meaningful drop that might, at first glance, suggest the fraud problem is gradually receding. The reality, however, is considerably more complicated. A falling overall fraud rate alongside a 180% surge in complex attacks suggests that the industry has become more effective at catching low-sophistication fraud while simultaneously being exposed to an entirely different tier of threat. The average is improving; the tail risk is worsening.
For financial institutions, fintechs, and digital platforms operating under tight regulatory scrutiny, this dynamic has direct implications for how they architect their risk frameworks. Regulators across major jurisdictions have consistently communicated that know-your-customer (KYC) and anti-money-laundering (AML) compliance cannot be treated as static checkbox exercises. The European Banking Authority and the Financial Action Task Force have each signaled in recent guidance cycles that adaptive, risk-based approaches to customer due diligence are not merely best practice — they are an expectation. A risk-classification engine that dynamically routes applicants based on real onboarding signals aligns closely with this regulatory posture.
The product is framed as a starter kit, which carries important strategic implications for how Sumsub is positioning it in the market. Rather than pitching a fully integrated enterprise overhaul, the company is offering an accessible entry point — a modular tool that organizations with varying levels of technical maturity can deploy without restructuring their existing compliance infrastructure. This approach acknowledges that not every institution has the internal engineering bandwidth or budget to build bespoke risk-scoring systems. For smaller fintechs and emerging-market operators in particular, a ready-made classification layer that plugs into existing onboarding flows could meaningfully compress the time-to-competency gap in fraud prevention.
The broader context for this launch is a fraud ecosystem that has been materially reshaped by the proliferation of generative artificial intelligence tools. Synthetic faces, cloned voices, and fabricated identity documents have become far more accessible to bad actors than they were even two years ago. The 180% rise in complex attacks almost certainly reflects, at least in part, the democratization of these capabilities. Fighting AI-assisted fraud with manual review alone is no longer viable at any meaningful scale, which makes automated risk-tiering not simply a convenience but a structural necessity.
What This Means for the Industry
Sumsub's fraud prevention starter kit arrives at a moment when the gap between the fraud rate headline and the fraud complexity reality has never been wider. The decline from 2.6% to 2.2% in global identity fraud rates is a genuine improvement worth acknowledging — but institutions that allow that number to breed complacency do so at considerable risk. A 180% surge in complex attacks means that the incidents that do get through are increasingly difficult to detect, increasingly damaging in their impact, and increasingly expensive to remediate. The case for structured, scalable risk classification at the point of onboarding has never been stronger, and Sumsub's decision to package that capability into an accessible starter format lowers the barrier to adoption for exactly the institutions that need it most.
Written by the editorial team — independent journalism powered by Codego Press.