FB pixel

US agencies not complying with AI laws and rules; maybe a bot would help

US agencies not complying with AI laws and rules; maybe a bot would help
 

An audit of AI implementations by major departments of the U.S. government has found that not all planned, in-production or in-use algorithms adhere to White House safety and transparency requirements.

The audit was conducted by the Congress’ Government Accountability Office, which examines operations of the legislative branch. It suggested 35 recommendations.

GAO officials looked at current and planned uses of AI by 23 agencies. They also wanted to see how accurate and complete AI reporting is and how closely agencies complied with AI laws, policies and guidance.

Of the 23 agencies asked to inventory their AI use cases (in production and in use), only three said they have no use for AI. The other 20 reported 1,200 current and planned AI use cases. Of those, 200 algorithms were in use.

While the audit was interested in AI activity in general, a number of the agencies and departments, including NASA, State and Health and Human Services, are involved with biometric systems.

The Department of Homeland Security has deployed facial recognition and is developing more code. NASA and the Commerce Department had the most AI use cases.

A top concern for GAO is that there is no common federal definition for AI. That was apparent when it was learned that projects by agencies did not involve AI.

Meeting all recommendations would put agencies in compliance, according to the GAO’s report.

Fifteen agencies need to bring their AI use case inventories up to date by including information required to be reported and aligning data with government guidance.

Other agencies that interact with all other agencies (including the Office of Management and Budget) need to be more active in issuing guidance.

And 12 agencies have to catch up with their responsibilities. They have to implement AI governance requirements, including creating a plan for conducting annual inventory updates.

GAO has previously found that agencies were using facial recognition from non-federal partners without knowing exactly who those partners were, and has urged DHS to implement its recommendations on biometrics.

Article Topics

 |   |   |   | 

Latest Biometrics News

 

Report demystifies India’s unique face biometrics market beyond the benchmarks

Biometric authentication is taking off in India as the country’s government and market align around identity as a trust layer…

 

Trust inevitable in building human rights-sensitive digital ID systems

Some digital rights advocates who spoke at the recent ID4Africa 2026 AGM in Abidjan emphasized that for African governments to…

 

Nepalese raise concerns over new DPI loans amid previous project failures

Some experts have expressed apprehensions that the government of Nepal has contracted a new loan for the implementation of a…

 

GripID introduces ultra-compact multimodal biometric enrollment kit

France-based GripID has unveiled the compact V10 multimodal biometric enrollment kit for registration to national ID and civil digital identity…

 

Australia opens feedback on verifiable credential policy, trust framework proposals

Australia’s Department of Finance is inviting community feedback on a policy for using verifiable credentials proposed by the Commonwealth. The…

 

FBI warning on Kali365 phishing kit exposes limits of weaker authentication

A new Federal Bureau of Investigation (FBI) warning about a phishing-as-a-service kit targeting Microsoft 365 accounts is underscoring why major…

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Biometric Market Analysis and Buyer's Guides

Most Viewed This Week

Featured Company

Biometrics Insight, Opinion

Digital ID In-Depth

Biometrics White Papers

Biometrics Events