FRISS for Insurance Fraud Detection and Underwriting Risk

In-Depth Review

FRISS was founded in Utrecht, Netherlands in 2006, making it one of the older pure-play insurance fraud detection vendors. The company has built its platform exclusively for insurance, with no cross-industry ambitions, and has accumulated over 300 implementations across 40+ countries. That single-industry focus shows in the product: the scoring models, the entity linking logic, and the case management workflows are all designed around insurance-specific fraud scenarios rather than adapted from generic financial crime detection.

What FRISS Does for Insurance Operations

The core concept is the Trust Automation Score, a 0-100 risk rating applied to every policy application and every claim at intake. At the underwriting stage, the score flags applications with misrepresentation risk, identity concerns, or patterns associated with application fraud. At FNOL, the same scoring framework flags claims that warrant closer investigation before they enter the standard adjustment workflow.

The score itself is deliberately simple. FRISS designed it so that underwriters and adjusters can act on it without data science background: a high score means review it, a low score means process normally. The underlying models are more sophisticated, but the output is designed for frontline staff, not analysts.

Network Detection

Where FRISS differentiates most clearly is network detection. The platform links entities (people, addresses, phone numbers, vehicles, providers, bank accounts) across the entire book of business and builds relationship graphs. This catches organized fraud that no single-transaction scoring model will find: a repair shop appearing repeatedly across unrelated claims, a medical provider billing consistently with the same attorney, or a cluster of applications from different names but the same IP address.

For carriers with large, diverse books, this cross-portfolio visibility is the primary value. First-party opportunistic fraud (a policyholder inflating one claim) is relatively easy to score. Organized provider fraud or staged accident rings require the network view.

Underwriting and Claims in One Platform

Most fraud detection vendors focus on either underwriting or claims. FRISS covers both, which matters because underwriting fraud and claims fraud are often linked. A policy obtained through misrepresentation is more likely to generate a fraudulent claim. By scoring both sides, FRISS can surface those connections and flag claims where the underlying policy was itself suspect.

Pricing and Deployment

FRISS does not publish pricing. Contracts go through enterprise sales and typically involve multi-year commitments. Deployment timelines vary by integration complexity, but expect six to twelve months for a full implementation including model calibration on your historical data. The platform requires access to your claims and policy data to train and validate the scoring models for your specific book.

Who This Is Best For

FRISS fits carriers running substantial claim volumes across multiple lines and, ideally, multiple geographies. Its multi-country deployment base is a genuine advantage for international operations. For North American carriers operating domestically only, validate that the fraud pattern coverage is sufficiently localized before committing. The product’s European heritage means it is strongest in continental and UK markets, though North American adoption has grown.