IRS Paid Palantir $1.8M to Build AI Tool for Targeting Tax Audits
The IRS paid Palantir $1.8 million last year to develop a custom tool called SNAP that helps the tax agency identify "highest-value" targets for audits, tax collection, and criminal investigations from its maze of legacy systems.
The Problem
The IRS currently uses:
- 100+ business systems built over decades
- 700 methods for selecting audit and investigation targets
- Highly fragmented, inefficient case selection
The agency itself acknowledged: "This fragmented landscape can lead to duplication of effort and cost, poor understanding of gaps in the coverage, and suboptimal case selection."
The Solution: SNAP
Palantir's "Selection and Analytic Platform" (SNAP) is designed to:
- Streamline identification of potential fraud cases
- Surface "key information about contracts, vehicles and vendors" from unstructured data
- Help human auditors identify red flags they might otherwise miss
- Currently in pilot program phase
The Relationship
- The IRS has used Palantir technology since 2014
- Total Palantir contracts with the IRS exceed $200 million in awarded contracts and obligated payments
- The agency is now interested in deepening its relationship with Palantir
Concerns
The use of AI-powered surveillance tools for tax enforcement raises significant civil liberties questions:
- Algorithmic bias: AI tools may disproportionately flag certain demographics
- Transparency: How SNAP makes its decisions may not be publicly auditable
- Privacy: The tool aggregates data from multiple IRS databases and unstructured documents
- Scope creep: A pilot program could expand to broader surveillance
Palantir is known for its work with intelligence agencies and immigration enforcement (ICE), and its tools have been controversial for their role in government surveillance programs.
What's Next
The IRS asked Palantir to expand SNAP's capabilities, indicating the agency is moving toward broader adoption. Both Palantir and the IRS declined to comment on the specifics of the program.
This partnership signals a broader trend of government agencies turning to AI and big data analytics firms to modernize aging enforcement infrastructure.