Experian launches AI fraud detection tool for UK banks

Experian’s new tool is called Transaction Forensics, an AI‑powered fraud and AML solution developed with Resistant AI to help UK banks and other financial firms detect APP fraud, mule activity, and money laundering in real time.

What Experian launched

• Product name: Transaction Forensics, positioned as an AI‑driven fraud and anti‑money laundering platform for the UK financial services sector.

• Partners: Jointly developed by Experian and Prague‑based fraud analytics specialist Resistant AI, following Experian’s strategic investment in Resistant AI in 2025.

• Availability: Now live and available to UK businesses across the banking and wider financial services market.

• Payment rails covered: Supports Faster Payments, BACS and CHAPS bank‑to‑bank payments.

How the AI works

• Model stack: Uses more than 80 AI models combining behavioural analytics, transaction pattern analysis, and Experian’s proprietary consumer and commercial data.

• Data fusion: Fuses banks’ internal transaction data with Experian’s external credit, identity, fraud and AML datasets plus historical behavioural signals to infer the intentbehind each transaction.

• Real‑time performance: Exposes API‑based risk scores and insights in near‑real time; Experian markets response times at under 200 milliseconds for decisions.

• Deployment model: Designed to sit as an additional analytical layer on top of existing monitoring systems, or be targeted to specific higher‑risk transaction cohorts rather than fully replacing incumbents.

Fraud and crime types targeted

• Authorised push payment (APP) fraud: Focus on scams where customers are tricked into sending funds themselves; Transaction Forensics looks for behavioural and transaction anomalies that suggest social‑engineering patterns rather than simple rule breaches.

• Money mule networks: Detects mule accounts and mule herds by analysing flows of funds, counterparties, and behavioural histories over time.

• Money laundering and complex schemes: Assesses risk at transaction, customer and company level to identify layered or structured activity that might evade traditional rules‑based monitoring.

Example: A Faster Payment from a long‑standing low‑risk customer to a new beneficiary overseas at an unusually high value and at an unusual time could be flagged because the behavioural pattern deviates from the customer’s history and matches known scam or mule typologies.

Reported performance uplift

In pilots and early trials across UK financial institutions, Experian is highlighting three headline results:

• 200% increase in detected APP fraud compared with existing systems alone.

• 80% reduction in false positives, meaning far fewer legitimate transactions are incorrectly flagged.

• 50% reduction in overall alert volumes, helping investigation teams focus on genuine threats rather than noise.

Experian also emphasizes that combining identity data with behavioural and transaction analytics is, in their view, a “market‑first” approach in the UK for this kind of cross‑rail, APP‑focused solution.

Positioning for UK banks

• Regulatory alignment: Aims to address increasing regulatory pressure in the UK around APP fraud reimbursement, AML controls, and explainable AI in risk decisions.

• Customer experience: By cutting false positives and alerts, banks can apply stricter fraud controls while maintaining relatively low friction for legitimate customers.

• Integration: Offered as an API‑enabled service that can plug into existing fraud stacks and payment workflows, rather than requiring a wholesale rip‑and‑replace.

For your purposes, are you mainly interested in this from a regulatory‑risk angle (e.g., APP reimbursement, explainability) or from a vendor/partnership perspective?