LexisNexis Risk Solutions Ships Self-Calibrating Fraud Engine

LexisNexis Risk Solutions has released Emailage Adaptive, a fraud-scoring engine that recalibrates itself rather than waiting for human tuning. The product blends an organisation's own transaction records, verified fraud cases and broader industry intelligence, updating its models as criminal behaviour shifts.

The launch was announced on 22 September 2026. Kimberly Sutherland, who oversees fraud and identity work globally at the firm, framed it as a counter to generative AI, which in her view lets attackers move faster than hand-tuned defences can. Dell Technologies is already running the system, according to the vendor.

Emailage Adaptive weighs identifiers such as email handles, device IPs, phone numbers and street addresses when generating a risk score, and it covers onboarding, ongoing account servicing and payment flows. In-house testing reported by the vendor flagged roughly nine in ten fraudulent cases inside its highest-risk bucket while cutting false alarms by over 80%. Dell, meanwhile, is said to have doubled how much fraud it caught among high-risk transactions and trimmed manual checks of low-risk ones by 83%. Reason codes accompany every score, and the company says explainability, privacy and bias controls are built in.

Fraud teams have long juggled stale rules against adversaries who retool constantly; a model that relearns on its own closes that window. If the figures hold, vendors without self-updating engines will feel the squeeze, and buyers will press harder on how much analyst time a fraud stack really eats.

Why it matters for Singapore: Banks and fintechs here sit on one of Asia's busiest cross-border payment corridors, so fewer false alarms translates directly into freed-up compliance headcount and quicker onboarding. A tool that ships its reasoning alongside each score also suits the trust-and-explainability framing Singapore keeps attaching to its AI ambitions, giving local adopters an easier internal case to make.