Operationalizing AI for public sector fraud prevention

Summary

Public-sector agencies face fraud methods including synthetic identities, deepfake-enhanced documentation, and personalized social engineering, while legacy controls remain fragmented and difficult to scale. The fictional Services Bureau demonstrates an operating model that combines Databricks Apps, Lakebase, Unity Catalog, Delta Sharing, Agent Bricks, and AI/BI Genie in a single fraud-operations environment. Governed data lands in Delta tables; Unity Catalog applies attribute-based access control, masks PII by role, and provides lineage, while agents connect live lakehouse queries, agency policies, and external fraud signals through MCP. Analysts review evidence and recommendations, then approve, override, or escalate cases, keeping human judgment central. The described workflow turns weeks of manual investigation into a day, supports dashboards and conversational SQL, and is presented as making fraud decisions faster, more secure, transparent, and defensible.

Context

Public-sector fraud controls are described as fragmented, time-consuming, and difficult to scale. Agencies face synthetic identities, deepfake-enhanced documentation, personalized social engineering, and growing fraud volumes, creating a need to connect trusted data, intelligence, governance, and operational workflows.

Approach / What changed

The fictional Services Bureau uses Databricks Apps powered by Lakebase to combine fraud cases, governed evidence, dashboards, and embedded agents. Delta tables hold incoming data; Unity Catalog applies attribute-based access control, PII masking, and lineage. Agent Bricks coordinates Genie, a policy-grounded Knowledge Assistant, and an external MCP server, while analysts retain authority to approve, override, or escalate recommendations.

Takeaways

• Unity Catalog uses attribute-based access control and tagged masking policies so access to PII varies by role, while table- and column-level lineage shows where data originated and flows downstream.
• Agent Bricks coordinates Genie for live lakehouse statistics, a Knowledge Assistant for agency procedures and citations, and a web agent using an external MCP server for emerging fraud signals.
• The workflow keeps analysts in the loop: they review evidence and rationale before approving, overriding, or escalating recommendations, with labeling sessions available to refine agent outputs.

Related reading