US housing department plans AI checks on how $77 billion in grants is spent
HUD's finance chief wants an invoice behind every payment a grantee makes. Housing advocates worry about errors and the burden on small organisations.
The US Department of Housing and Urban Development (HUD) plans to launch an AI tool to analyse how grant recipients spend their money, FedScoop reported. Irving Dennis, acting chief financial officer since January 2025, told the Senate Banking Committee at his nomination hearing that the department does not currently know how grantees spend the roughly $77 billion it sends out each year.
"So what I'm putting in is using AI and technology to get information for every check that's cut from the grantee to the vendor," Dennis said. "I want the invoice or the contract that supports that vendor. We'll get that using AI." He described a pilot with 10 grantees as a "great success" and said HUD was working towards a launch on 30 September.
Dennis said grantees could submit data daily, weekly, monthly or quarterly: "I don't want this to be a burden for the grantees."
Concerns
No matching use case appears in HUD's published AI inventories, though a source told FedScoop the list is being updated. Housing and technology experts said there could be efficiency gains, but raised questions about whether automated reviews could add uncertainty to grant delivery, disadvantage small grantees without the staff to contest AI errors, and open the door to politically motivated decisions.
Tracing money without stopping it
Following grant money to the final vendor is a real gap in many governments, and fraud in housing and disaster programmes is costly. The design questions are the ones any agency should ask before automating oversight: what triggers a flag, who reviews it, how a grantee can respond and how quickly payments resume when the model is wrong. For organisations housing vulnerable people, a frozen payment is not a neutral outcome.
Newsroom, Online
Stories by the IQGov newsroom are researched and written by AI agents to our editorial rulebook, checked against their sources by a separate fact-checking agent, and approved by the publisher before publication. Every fact links to its source.
