Modern IDV and fraud prevention is a fortress of orchestrated layers

In the story of the Three Little Pigs, the Big Bad Wolf is famously unable to blow down the brick house favored by the third and smartest pig. But consider a more persistent wolf, who returns with a jackhammer. The logical progression would eventually see the pigs realize that, no matter what it is made of, one wall is not enough – and proceed with the construction of ramparts, a moat, a barbican with portcullis, battlements with towers, and possibly an inner sanctum outfitted with a panic room and a gun safe.

In other words, a layered approach to defense.

The same phenomenon can be observed in today’s evolving biometrics and fraud prevention sector. As various threats mutate and conjoin to maximize damage, providers are realizing that selling bricks for a wall is less helpful than providing a complete, responsive defensive strategy which factors in all of the potential risks, from deepfake and synthetic media to injection tactics. The value is increasingly in orchestration: designing and maintaining an effective security posture that can adapt to ongoing advances in fraud.

Core elements must work together

This recognition is beginning to show itself in the market, as providers fashion their products as engines for secure identity verification and fraud detection across the customer journey.

A recent article in Technology.org applies the layer theory to the verification workflow. Layer one, it says, is reading and validating the document to confirm it is genuine. “Modern AI-powered ID verification starts with computer vision. Optical character recognition extracts the text, while machine-readable zone parsing decodes the encoded strip on passports and ID cards.”

“The software then checks the document against known templates for more than 15,000 identity types across 250-plus countries, looking for the security features, fonts, layouts and holographic cues that a forgery tends to get wrong.”

Layer two does the same for a human. Is the person submitting the document real and present? “This is where biometrics enter. The system captures a selfie and performs facial matching against the photo on the document, measuring the geometric relationships between facial features rather than comparing images pixel for pixel.”

But “the critical piece is liveness detection, and it is the front line against deepfakes.”

“Because generative models keep improving, this layer has become an ongoing research problem: detectors are retrained continuously as new attack methods appear, and the best systems treat liveness as a moving target rather than a solved feature.”

The third layer is continuous monitoring of “signals that have nothing to do with the ID document at all” – a transaction analysis that weighs amounts, locations, velocity and timing against a user’s history.

“Behavioral biometrics analyze how a person types, moves a mouse and navigates a session, building a profile of normal interaction that is remarkably hard for an impostor to imitate. Device intelligence fingerprints the hardware and flags mismatches, new devices and the emulators that fraud rings favor.”

The kernel of the matter is that, for modern fraud prevention, no individual technology is sufficient on its own. “A deepfake might fool a face match but fail a passive liveness test. A stolen identity might pass document checks but fail behavioral analysis.” And the implication for the industry is that “verification and fraud prevention are no longer separate products but a continuous pipeline that runs from the first document scan through every subsequent transaction.”

Protection, simplification exist in delicate tension

The problem is, at the same time that fraud risks are driving the need for layered trust, the identity verification landscape is becoming increasingly fragmented. So says LSEG Risk Intelligence in announcing its Identity Gateway integration tool for identity verification.

“Organizations often rely on multiple identity verification APIs, providers and schemes across different markets, creating inconsistent experiences and operational overheads,” it says. The idea is to offer a single integration layer for multiple IDV APIs, to simplify the experience – in other words, to provide orchestration.

“Identity Gateway operates as an orchestration layer between customer journeys and trusted identity providers,” it says. When a user needs to verify their identity, “it routes verification requests to the appropriate scheme and returns consistent, structured identity data that can be integrated into existing systems.”

LSEG says the model reduces technical complexity, makes for a better user experience, and reduces time to market for new regions.

Consolidation of services toward real-time trust

So goes the slow waltz toward consolidation, and a business model that builds out from a product’s purpose and its users’ needs. It makes for slippery terminology and regulatory complexity – but a static approach to identity verification and fraud prevention is no longer viable.

Ask the British Council, which has enlisted Daon to provide identity verification services across its global testing and digital education portfolio. Per a release, the deployment will see Daon’s TrustX platform “used to verify the identities of test takers at multiple points throughout the testing journey, supporting a more secure and consistent approach to identity assurance.”

“Organizations like the British Council operate at a scale where identity is more than just a security function,” says Tom Grissen, CEO of Daon. “It’s what underlines trust in the institution and the services it provides. This deployment reflects something we’re seeing across multiple sectors, where identity verification is becoming an ongoing, integrated part of the user journey rather than a single, static checkpoint. Platforms like TrustX are designed to enable organizations to orchestrate identity across channels, use cases, and geographies without adding friction for users.”

That nicely sums up the current market sweet spot: two-way integrations between IDV and fraud prevention that cover the breadth of protections for both. Stablecoin-focused fraud and risk intelligence firm Sardine recently announced a partnership with Infinite, integrating its tech into the latter’s compliance platform and account infrastructure. Identity verification provider SQR has teamed up with Proofdesk, the Isle of Man’s dedicated AML/CFT compliance platform, to service the local market; per a release, “by connecting the two products, firms will be able to trigger and manage identity checks from the same platform that they use for structured reviews and customer risk assessments.”

Beware the wolf in deepfake clothes

If the goal is to be comprehensive in offering digital identity services – biometric verification, document checks, liveness detection, deepfake detection, injection attack detection, privacy preserving age assurance – something has to serve as the glue holding it together and giving it structure. This is presumably why so many biometrics and IDV providers swear by AI or machine learning as a core pillar of their business. It is what moves all the pieces into place.

As Technology.org puts it, “the threats change weekly, so static rules age badly. Models that learn from new fraud patterns, retrain on fresh attack data and adjust risk thresholds automatically are the only way to keep pace with adversaries who now have AI of their own.”

Staying one step ahead is one thing – a matter of pace. But it requires a recalibration of how organizations understand identity verification and trust. “The organizations that treat it as living infrastructure rather than a fixed feature,” says the article, “are the ones that will keep trust intact as the fakes get better.”

The wolf never stopped trying; he just changed his clothes – and his arsenal. Today, that includes the same AI technologies available to any enterprising criminal. Perhaps the wolf no longer has a wolf’s face, but that of a lamb. Perhaps he sounds like Grandma. He says that he will never die; the pigs would do well to listen. By now, one hopes their castle is managed by an autonomous security system that can shift shapes just as well as the wolf can.

Article Topics

biometric liveness detection  |  biometrics  |  Daon  |  deepfake detection  |  digital identity  |  identity orchestration  |  identity verification  |  LSEG Risk Intelligence  |  Sardine  |  SQR Group