The Trump administration apparently still hasn’t learned its lesson about double-checking sources.
A federal labor agency flagged so many problems in a recent arbitration filing that it suggested the Department of Health and Human Services used artificial intelligence to draft its brief.
In a section where HHS was arguing that an arbitrator got key facts wrong, one of the government’s “supporting legal authorities has a fictitious citation,” a Federal Labor Relations Authority decision says.
“The Agency cites ‘U.S. Dep’t of the Air Force, 913th Air Wing, 57 FLRA 469, 472 (2001),’ which does not exist,” the decision says in a footnote.
Trending
It was one of 10 “entirely fictitious” citations the agency says it found. And overall, 22 of the HHS filing’s 29 footnotes — and 35 of the 42 total legal or record citations — “are fictitious, erroneous, or irrelevant.”
This is not the first time HHS has had trouble with citations in the second Trump administration. Last year, The Washington Sun found the department inaccurately cited more than 500 scientific sources to back up its “Make American Healthy Again” commission report — in some cases even making up studies.
In this latest incident, multiple times the decision points out where HHS attorneys cited a case that refers to an irrelevant part of the law. There are also incorrect page numbers, case numbers and other details.
“Based on the nature and scope of these errors, it appears that the Agency may have used a generative artificial-intelligence (AI) tool in drafting its brief,” the decision says.
HHS was asking the FLRA to reverse a decision from a third-party arbitrator who sided with a labor union in a challenge to a presidential memorandum sending federal workers back to in-person work.
HHS not only didn’t get what they wanted, but it also got a lecture from FLRA about using AI.
“While nothing prohibits parties from using AI tools when preparing filings for submission to the Authority, we caution those who do so to be aware that AI continues to present significant risks of error,” the FLRA wrote.
“Therefore, parties should take steps to verify the accuracy of any AI-generated material and ensure that their filings satisfy all regulatory requirements.”
The FLRA used its own footnote there, pointing to a 2024 guidance on the use of AI-based tools in court “on how to follow filing rules when drafting documents using AI, including reviewing and correcting AI-generated text.”
The HHS Office of the General Counsel did not respond to a request for comment about whether it used AI to write its filing.
And the issue has popped up elsewhere as AI use has become more widespread. In July, a federal judge cautioned an attorney with the Justice Department about AI use that led to inaccurate citations in an immigration case.
President Donald Trump has embraced artificial intelligence, and multiple agencies across the federal government have confirmed they are using the technology.
Author
