INETCO Launches Agentic AI Module BullzAI Investigate, Cutting Fraud Case Review Time to About 20 Seconds

Canadian payment fraud prevention company INETCO launched an agentic AI module, BullzAI Investigate, on August 3, 2026, to accelerate fraud case reviews for banks and payment service providers. The module is available now.

The module runs on a fully on-premises proprietary small language model, meaning transaction data never leaves customer infrastructure. The AI agent aggregates transaction data, identifies behavioral patterns, and automatically triages alerts, forwarding the highest-risk cases to analysts. Each recommendation is accompanied by an explainable risk score, which INETCO says provides investigators with an auditable basis for decisions. The system uses a human-in-the-loop supervised learning mechanism, absorbing analyst feedback after cases are closed.

According to the company's early data, case review times drop from 10–30 minutes to roughly 20 seconds, a 97–99% reduction. Recommendation precision is approximately 95%, although the statistical methodology behind this figure has not been independently published. INETCO CTO Ugan Naidoo said that agentic AI handles data aggregation and triage, but humans retain oversight and final decision authority. The company also cited a Chartis Research report on an unnamed South African bank's deployment of BullzAI as third-party validation, but did not disclose the research methodology or deployment scale.

Source: https://thefintechtimes.com/inetco-adds-agentic-ai-module-to-cut-fraud-case-review-times/