HUD Withholds 100-Plus DOGE AI Records, Citing Nonexistent FOIA Exemptions
Updated
Updated · WIRED · Jul 14
HUD Withholds 100-Plus DOGE AI Records, Citing Nonexistent FOIA Exemptions
1 articles · Updated · WIRED · Jul 14
Summary
More than 100 HUD records sought by Democracy Forward were withheld, obscuring how DOGE staff used AI to shape housing policy and regulatory decisions.
Several denials cited "deliberative AI input," "draft of AI prompt," or "deliberation of AI prompt"—grounds privacy experts say do not exist under FOIA, even though most records were also tagged under Exemption 5.
Named but unreleased files such as "GPT defined Econ Analysis approach 11 10 25.docx" and "RegulatoryAnalysisPrompt.pdf" suggest DOGE officials were building prompts and analyses for regulatory review inside HUD.
Christopher Sweet and Scott Langmack were previously tied to DOGE's HUD effort to flag rules for rescission; Langmack now leads deregulation AI at OMB.
The dispute highlights a wider gap in U.S. law: agencies generally do not have to disclose when AI helps draft rules or policy, even as critics warn the public cannot assess bias or errors without seeing the prompts.
As AI helps draft federal regulations, what prevents algorithmic bias from becoming the law of the land?
When courts rule AI chats aren't private, why does a federal agency claim they are privileged government secrets?
HUD’s AI Secrecy Sparks Legal Battle: The Fight Over “AI Privilege,” FOIA, and the Future of Government Transparency
Overview
The Department of Housing and Urban Development (HUD) faces growing controversy over its secretive use of artificial intelligence, as it withholds key documents and refuses to disclose details about its AI systems. By citing reasons like 'deliberation of regulatory changes' and 'presidential communications privilege,' HUD has sparked legal and public debate about whether it is creating a new kind of 'AI privilege' to shield its operations from scrutiny. This lack of transparency, combined with reports of internal challenges and unanswered questions about personnel, highlights serious concerns about accountability and the influence of AI on public policy.