HUD Withholds Records on DOGE's AI Use in Housing Policy

The Department of Housing and Urban Development has withheld more than 100 documents about how members of the so-called Department of Government Efficiency used artificial intelligence to shape policy decisions, according to records obtained by Democracy Forward, a nonprofit legal organization.

Among the justifications HUD cited for denying the Freedom of Information Act requests: a privilege for AI that does not exist in law.

How DOGE deployed AI at HUD

According to WIRED, which first reported these details, Christopher Sweet and Scott Langmack worked as part of the DOGE team at HUD last year. Sweet, then a third-year economics student at the University of Chicago, focused on using AI to identify agency rules for potential rescission or contract cancellations. HUD employees told WIRED at the time that staff were asked to provide feedback on regulations flagged by the AI system.

Sweet has since graduated. Langmack now serves as executive director of deregulation AI at the Office of Management and Budget, according to his LinkedIn profile.

Documents whose titles were disclosed but contents withheld suggest the team used AI tools for policy analysis. One file labeled "GPT defined Econ Analysis approach" was exempted as "deliberative AI input." Another titled "RegulatoryAnalysisPrompt.pdf" indicates the team explored creating prompts for regulatory analysis.

The transparency problem

Nearly all withheld documents were denied under FOIA Exemption 5, which protects "deliberative process privilege"—material where federal workers debate policy before decisions are made. The privilege exists to encourage candor among government employees.

But legal experts say the privilege does not extend to interactions between humans and AI systems.

"There is no AI exemption under FOIA," John Davisson, deputy director of enforcement at the Electronic Privacy Information Center, told WIRED. "AI systems, computers are not entitled to candor."

Some documents were also withheld citing "presidential communications privilege," which typically applies only to the president and immediate advisers—raising questions about the origin of certain AI prompts.

Why it matters

The lack of transparency around AI use in policymaking creates accountability gaps at a time when these tools are known to hallucinate, exhibit bias, or produce incorrect outputs. Without disclosure requirements, the public cannot assess whether AI-generated analysis influenced consequential regulatory decisions affecting housing policy. The invention of novel legal privileges to shield AI records from scrutiny sets a concerning precedent as government agencies increasingly adopt these technologies.

"If the government is going to use AI in formulating policy that affects us all, the public has a right to understand its impact," Dan McGrath, senior oversight counsel at Democracy Forward, told WIRED.

Currently, no U.S. laws require the government to disclose when AI has been used to create rules, policies, or regulations. HUD, OMB, and the White House did not respond to requests for comment.

These details were first reported by WIRED.