Australia's Insurers Launch Cross-Carrier AI Fraud Detection Platform with EXL and Shift Technology

Australia's insurance industry launched a national cross-carrier AI fraud detection platform, allowing competing insurers to securely share fraud patterns and coordinate investigations in real time—closing a structural gap that individual carriers cannot close on their own. The Insurance Council of Australia (ICA) partnered with analytics firm EXL and AI fraud specialist Shift Technology to build and operate the system, beginning with auto claims.

Why it matters: Insurance fraud isn't committed against one carrier at a time. Fraudsters exploit the fact that insurers don't share data. Cross-carrier AI detection closes that gap in a way that no individual carrier's fraud team ever could.

The Structural Problem With Single-Carrier Detection

A fraud ring submitting staged accident claims doesn't hit one insurer—it distributes submissions across multiple carriers simultaneously, keeping the per-carrier volume low enough to evade individual detection thresholds. Each insurer sees noise; the pattern is only visible when viewed across the industry.

This is the structural problem the ICA-EXL-Shift platform addresses. By aggregating and analyzing claims data across participating carriers, the system identifies coordinated fraud patterns that are invisible to any single participant—according to the Insurance Council of Australia's launch announcement.

What the Platform Detects

The three core fraud categories the system targets:

• Synthetic identities: Fabricated policyholder identities assembled from real and manufactured data, designed to pass individual carrier identity verification
• Manipulated images: AI-altered or staged claim photographs submitted as evidence of damage or accidents that didn't occur as described
• Coordinated submission patterns: Organized fraud networks submitting similar claims across multiple carriers in ways that generate cross-carrier signals no single insurer's data would reveal

EXL brings data analytics and claims processing infrastructure to the collaboration; Shift Technology contributes fraud detection AI models already deployed across more than 150 insurers globally. The ICA provides the industry coordination network that makes cross-carrier data sharing operationally and legally feasible.

The Scale of the Problem

The ICA estimates that AI-enabled insurance fraud exposure runs at up to 20 times the rate seen in banking. That gap reflects structural differences between the two industries: insurance claims involve physical events that are harder to verify than financial transactions, documentation is easier to fabricate, and per-claim fraud amounts are large enough to attract organized criminal operations.

Deepfake-generated claim photographs and AI-synthesized accident documentation have materially accelerated fraud sophistication over the past two years. Earlier research tracked a 2,137% increase in deepfake-enabled insurance fraud—a figure that frames cross-carrier AI detection not as a competitive differentiator but as a baseline industry requirement.

What This Means for Carriers and Policyholders

For participating insurers: The platform converts fraud detection from an individual-carrier capability gap into an industry-level defense. Carriers that join gain access to cross-carrier signals their own data cannot generate, and can coordinate investigations in real time rather than discovering fraud patterns retrospectively through accumulated loss experience.

For policyholders: Reduced fraud losses reduce the underwriting pressure that drives premium increases. Auto insurance fraud is the highest-volume fraud category across most markets—starting here maximizes the near-term impact on loss ratios.

For data governance: Cross-carrier data sharing raises legitimate questions about policyholder information flowing through the platform. The ICA's announcement does not detail the data minimization architecture or privacy controls in place. Those specifics will matter for regulatory compliance under Australia's Privacy Act and for consumer trust as the platform scales beyond its auto claims launch.

What to Watch

The auto claims launch is a proof of concept for expansion into home, commercial property, and health insurance claims as detection models are validated across more complex fraud patterns. The ICA-led, vendor-implemented coordination model—industry body sets the governance framework, technology vendors build the detection layer—is one that insurance markets in the UK, Canada, and the US will evaluate closely as their own industries face the same AI-enabled fraud acceleration.

By Hector Herrera | Sources: Insurance Council of Australia