DOGE Used AI for Housing Policy. The Government Won’t Say How

2026-07-14 · Source: WIRED - Ai · Field: Government & Public Sector — Public Policy & Governance, Regulatory & Compliance, Digital Government & E-Government · Depth: Novice, medium

Summary

The Department of Housing and Urban Development (HUD) has withheld over 100 documents concerning the Department of Government Efficiency's (DOGE) use of artificial intelligence to inform policy decisions, following a Freedom of Information Act (FOIA) request by Democracy Forward. DOGE team members, including Christopher Sweet and Scott Langmack, reportedly used AI to identify agency rules for potential rescission. HUD cited a nonexistent "AI privilege" and a presidential communications privilege, typically reserved for the president and immediate advisors, to justify withholding documents like "GPT defined Econ Analysis approach 11 10 25.docx" and "RegulatoryAnalysisPrompt.pdf." Experts like Tori Noble from the Electronic Frontier Foundation express concern over this lack of transparency, citing AI's potential for hallucination, bias, or error, while John Davisson of the Electronic Privacy Information Center confirms no "AI exemption" exists under FOIA, asserting AI systems are not entitled to candor.

Key takeaway

For legal professionals and policy makers evaluating government transparency, HUD's assertion of a non-existent "AI privilege" to withhold documents on AI-informed policy decisions sets a concerning precedent. You should challenge such FOIA denials, as AI systems are not entitled to the same deliberative process protections as human officials. Demand disclosure of AI prompts and inputs to ensure accountability and mitigate risks like bias or error in government policy formulation.

Key insights

The government is withholding details on AI use in policy, citing non-existent or misapplied privileges, raising transparency concerns.

Principles

• No "AI privilege" exists under FOIA.
• AI systems are not entitled to deliberative candor.
• Transparency builds public confidence in government AI.

In practice

• Request AI prompts to understand policy influence.
• Scrutinize FOIA denials citing "AI privilege."

Topics

• Government AI
• FOIA Transparency
• AI Policy
• Deliberative Process Privilege
• Public Records Law
• Regulatory Analysis

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Legal Professional, Policy Maker, AI Ethicist

Related on AIssential

• The complaint by Just Futures Law is formally a FOIA case, but substantively it challenges the secrecy around ICE, DHS, DOGE and Palantir tools. — Opaque data-fusion systems enabling coercive state action demand transparency for public accountability and legal assessment.
• Of Slop and Swarms: The First Amendment's Next Test — First Amendment protection for AI outputs should hinge on demonstrable human expressive contribution and output predictability.
• The End of the Fine Print: How AI is Democratizing the Right to Housing in the U.S. — AI tools can democratize legal information, empowering tenants to understand complex rental contracts.
• Anthropic and the Pentagon are reportedly arguing over Claude usage — AI companies face pressure from the U.S. military to permit broad use of their models, sparking policy disputes.
• AI as “Arbitrary” Intelligence — AI's "black box" and third-party data complicate judicial review of agency decisions, risking arbitrary outcomes and constitutional violations.
• » Detroiters cautious about government use of AI, U-M survey finds - Poverty Solutions — Public support for government AI applications is conditional, varying by use case and significantly impacted by data privacy concerns.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by WIRED - Ai.