Unemployment Insurance Fraud Detection and Prevention
Situation
Our client was a large state workforce agency, responsible for workforce development, job placement, rehabilitation, and unemployment insurance. During the early phases of COVID-19, the agency experienced a 10x increase of claims. This led to more federal funding and changes to processes, which together caused a surge in fraudulent claims. The agency urgently needed to enhance its fraud detection processes to manage the increased volume of claims, while still protecting the people supporting legitimate claimants.
Solutions Overview
Launched an AI/ML pilot program to eliminate one type of fraud.
Enabled victim notifications & prioritized high-risk cases for investigators.
Built automated workflows to streamline fraud detection.
Results
Fraud Detection Impact
The initiatives of our data science model was a part of identified approximately $3.8 billion in unemployment insurance fraud for the state, highlighting the effectiveness of our data-driven approach in preventing significant financial losses.
Verified Outcomes
Roughly $4B averted in fraud loss through consulting and collaboration, a 10x per capita gap between Texas and similarly sized states.
Sustained Capability
The client team is now able to independently evolve the system, and solve new problems with AI techniques.