LexisNexis self-calibrating fraud AI cuts false positives by 80%

Emailage Adaptive builds institution-specific fraud models that recalibrate automatically, as generative AI accelerates attack sophistication.

By Disrupts Editorial Team Mon, 28 September 2026 AI level 2 of 5

AI level 2 of 5: drafted by our AI editorial assistant from the company's announcement; fact-checked, edited and signed off by Mark Walker, Editorial Director. What the levels mean

A long data center aisle is flanked by dark server racks displaying numerous glowing orange and green lights, leading to a bright windowed control area with desks, chairs, and monitors, under bright overhead ceiling lights.

LexisNexis Risk Solutions has launched Emailage Adaptive, a self-calibrating fraud detection product that builds bespoke AI models for each client institution and updates them continuously without manual intervention. The release, timed to the company's internal pilot data from Dell Technologies, signals a broader shift in financial crime prevention: static, rules-based fraud engines are giving way to inference models that learn in near-real time from evolving attack patterns.

The product sits at the intersection of identity intelligence, payment security, and enterprise AI operations, making it a cross-sector story well beyond a single fintech product launch.

What Emailage Adaptive actually does

Rather than applying a single, industry-wide fraud model, Emailage Adaptive ingests each organisation's own transaction history, fraud outcomes, and industry signals to construct a tailored risk-scoring engine. That engine then recalibrates automatically as new confirmed fraud patterns emerge, producing what LexisNexis calls "explainable risk scores" backed by clear reason codes indicating which signals drove a particular decision.

The technical architecture draws on email address metadata, IP data, phone signals, and physical address history. LexisNexis points to the scale of the email-account universe as a key differentiator: with 7.9 billion email accounts globally, and roughly one third of users retaining the same address for more than a decade, longitudinal email-behaviour analysis provides a fraud signal that is both persistent and difficult to spoof cheaply.

In testing, the system reportedly captured around 90% of frauds among the highest-risk transactions while cutting false positives by more than 80%, according to the company. Dell Technologies, cited as the lead reference customer, says it has doubled its high-risk fraud capture rate and reduced manual reviews of low-risk transactions by 83% since adoption.

"The self-calibrating AI model has helped us modernise fraud prevention by minimising reliance on static rules and manual policy adjustments," said Jeremy Cole, director of business operations at Dell Technologies. "Our fraud team can now calibrate less, act faster and approve more legitimate transactions with confidence."

The generative AI arms race reshaping financial crime

The launch is explicitly framed against a deteriorating threat environment. LexisNexis' own Cybercrime Report 2026 finds that one in every eleven new account creations globally is a fraudulent attempt, a figure that points to the industrialisation of account-fraud operations. Kimberly Sutherland, global head of fraud and identity at LexisNexis Risk Solutions, argues that generative AI has accelerated attack velocity beyond the threshold that manual prevention strategies can sustain, creating a structural lag between when new fraud patterns emerge and when legacy defences respond.

That dynamic has macro-level implications. For cross-sector leaders, the arms race between generative-AI-enabled fraud and adaptive AI defences is no longer a specialist compliance problem: it sits at the heart of digital commerce infrastructure. Every sector processing high-volume digital transactions at scale, from retail e-commerce to travel booking to payments, faces rising exposure as attack tooling becomes cheaper and more accessible via commoditised large language models.

The capital landscape reflects this urgency. Enterprise fraud-prevention and identity-verification platforms have attracted sustained investment in recent years, with incumbents such as LexisNexis Risk Solutions (part of RELX), Experian, and FICO competing alongside specialist challengers. The differentiating battleground has shifted from data coverage, where incumbents hold structural advantages, to model adaptability and explainability. The latter matters because regulators in the EU and UK are increasingly requiring that automated credit and fraud decisions be explainable to consumers and auditable by compliance teams, a requirement that pure black-box models cannot easily satisfy. Emailage Adaptive's stated emphasis on reason codes and responsible-AI principles appears designed in part to anticipate that regulatory direction.

For investors tracking the convergence of AI infrastructure spend with financial services compliance, the broader question is whether the recalibration cycle these products promise can outpace a generative-AI-enabled adversary that can itself iterate rapidly. The answer will determine whether adaptive fraud models become a durable competitive moat or simply the next layer in an escalating cycle of measure and countermeasure.