Stop fraud before it drains your budget: A guide to AI-driven detection

Fraud is draining public sector budgets at a staggering scale. US federal improper payments hit $186 billion in fiscal year 2025 alone,¹ and UK government fraud and error runs £55 billion to £81 billion a year.² Every dollar lost is a dollar taken from citizens, patients, and students who depend on these programs.

Generative AI is making fraud easier to scale, from synthetic identities to automated phishing. This guide explains how agencies can fight back by unifying fragmented data, from claims and transactions to case files and call records, into a single foundation for detection and investigation.

You'll see how the Elasticsearch Platform combines entity resolution, customizable detection rules, machine learning, and AI-driven alerts to help teams spot patterns instead of chasing one-off leads. The guide also includes a real-world example from the California Employment Development Department (EDD).

Get the guide to see how your agency can move from reactive investigation to proactive fraud prevention.

What you'll learn

• How to unify fragmented fraud data: See how ingesting structured and unstructured data into one platform eliminates blind spots caused by disconnected systems.
• How entity resolution catches what manual review misses: Learn how linking records helps teams surface hidden connections between claims.
• How to build custom detection rules for known fraud tactics: Learn how to codify patterns your team already recognizes into automated thresholds.
• How machine learning flags anomalies that don't fit a fixed pattern: Understand how unsupervised and supervised ML models monitor for activity that deviates from historical baselines across your data.
• How the California EDD achieved a 99% reduction in mean time to response: See how the agency secured 850 billion records and connected 3,000 servers using Elastic AI and ML capabilities.

Additional resources

• How state and local agencies can get ahead of fraud
• Detecting the undetectable: Building a fraud detection framework

Footnotes

1. US GAO, "Payment Integrity: Agencies' Estimated Improper Payments Increased to $186 Billion in Fiscal Year 2025," GAO-26-108694, April 2026.
2. UK National Audit Office estimate, cited in "Public Sector Fraud Authority Annual Report 2024–25," GOV.UK; UK Department for Work and Pensions.