TECHNOLOGY SUMMARY

FraudSphere is a multi-agent fraud intelligence system modeling the interdependent strategies of fraudsters, auditors, clients and regulators. Using Adversarial Risk Analysis and LLM-based generative simulation, it quantifies strategic dependencies and produces synthetic fraud scenarios for realistic red-teaming, advancing fraud detection from static analysis to adaptive, explainable, and continuously learning intelligence.

AREA/MATURITY/AWARDS

Primary Application Area: AI & Digital Transformation

Technology Development Status: Prototype

Technology Readiness Level: TRL 4

Vetted Programs/Awards: Relevant ideas are funded by AFOSR (2021-2024, NSF ExpandAI (2024-2026), 2020 BBVA Foundation

SHOWCASE SUMMARY

Organization Type: R&D/Laboratory/University

Website: https://faculty.txst.edu/profile/1238240

Showcase Booth: 414

MARKET KEYWORDS

Market Keywords: AI-Powered Fraud Detection, Risk Intelligence and Compliance Analytics, Synthetic Data and Red-Teaming