PRAC Deploys AI Fraud Prevention Engine
The US Pandemic Response Accountability Committee told Congress on Tuesday that its Fraud Prevention Engine, trained on roughly five million Small Business Administration Economic Injury Disaster Loan applications, could have flagged potentially tens of billions of dollars in fraudulent payments before disbursement. The AI, which processes about 20,000 applications per second and combines unsupervised, supervised, and rules-based components, has helped recover just over $500 million.
Key Points
- 1Processes 20,000 applications per second, trained on about five million SBA EIDL application records
- 2Combines unsupervised, supervised, and rules-based detection to uncover anomalies and linked fraudulent accounts
- 3Enables agencies to prioritize oversight and potentially prevent billions in improper pandemic payments
Scoring Rationale
Official, high-impact deployment potential with demonstrated recoveries, but limited novelty and retrospective training data constrain immediate transformative effect.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
