Incode has launched TrustGraph, a biometric-anchored, privacy-preserving trust network aimed at repeat and agentic fraud across institutions, the company said on 14 October 2026. Early-access testing showed up to 53% more fraud detected than existing models alone, Incode reported. The stack builds on peer-to-peer cryptographic technology from Identiq, which Incode acquired in June 2026.
Matching uses secure multi-party computation-style techniques without pooling or exchanging customer records and without a central customer data lake. Founder and chief executive Ricardo Amper said “Fraud is networked. Defenses are still siloed,” arguing attackers benefit when each institution sees only its own interactions. Interac’s Navjeet Grewal said the network lets institutions benefit from fraud signals “without exposing customer data.”
Incode projected agentic fraud could account for 60% of attempts by the end of 2026 — a company forecast, not an industry consensus figure. TrustGraph is positioned for financial institutions facing deepfakes, synthetic identities and romance scams that span multiple organisations.
Incode said TrustGraph is offered alongside its identity verification stack so institutions can combine onboarding assurance with cross-network repeat-fraud signals.
Because matching does not require a shared customer data lake, Incode argues regulated institutions can participate without creating a new centralised PII repository.