Rights groups urge US executive branch to keep gov’t agency AI inventory

The U.S. Office of Management and Budget should preserve existing AI transparency and safety protections as it acts on Executive Order 14179, the American Civil Liberties Union and a host of other advocacy groups say in an open letter.

The ACLU says Memorandum M-24-10’s provisions on inventories of government AI use cases and the minimum practices for AI uses that impact people’s rights and safety are too important to be set aside as the Trump Administration goes about “Removing Barriers to American Leadership in Artificial Intelligence.”

The new Executive Order comes along with a repeal of the Biden Administration’s EO 14110, but not the Memorandum, which followed from it.

A year ago, the World Privacy Forum called the Memorandum “a seismic policy shift” for biometrics providers and users.

AI is already playing a part in people’s lives, the ACLU argues, and the noted provisions are mandated by the Advancing American AI Act and the AI in Government Act of 2020. The group further says that President Trump’s Executive Order 13960, in directing agencies to build public trust and confidence in AI by protecting Americans’ rights and freedoms, laid the groundwork for Memorandum M-24-10.

ACLU identifies rules around risk assessments and facial recognition in law enforcement and replicating a person’s likeness of voice without consent as important rights protections built into the Memorandum.

Access Now, the Electronic Frontier Foundation, the Electronic Privacy Information Center (EPIC), New America’s Open Technology Institute and a pair of Christian groups are also among the 13 signatories to the letter.

M-24-10 also specifies that data used in AI should be assessed for bias, and disparities that might lead to bias mitigated. NIST’s assessments of demographic differentials predate the Memorandum, however, with a specific FRVT on demographic effects published in 2019, and FRTE 1:1 reports including demographic data all the way back to 2017.

A blog post from the Open Technology Institute argues that the AI inventory published by the Department of Homeland Services is particularly important to preserving public rights and safety.

DHS’ own 2024 AI use case inventory specifies 18 “rights-impacting cases” each in law enforcement and facial recognition among its AI deployments.

The Open Technology Institute calls DHS’ AI inventory “an essential element of public accountability.”

Article Topics

ACLU  |  AI  |  artificial intelligence  |  biometrics  |  DHS  |  regulation  |  U.S. AI policy  |  United States