- Timothy Burke is accused of grabbing unaired Fox News footage using someone else's credentials.
- A judge scolded his lawyers for misrepresenting and making up legal precedents in an attempt to get the case thrown out.
- A lawyer for Burke admitted to using ChatGPT and Westlaw's AI features without checking their output.
A lawyer for Timothy Burke, the journalist indicted over leaked Fox News footage, admitted in a court filing Monday that he used ChatGPT and other AI tools to write an error-filled legal brief.
Last week, Judge Kathryn Kimball Mizelle said a filing by Burke's lawyers contained "significant misrepresentations and misquotations" and demanded an explanation. On Monday, the lawyers, Michael Maddux and Mark Rasch, said the errors happened because of Rasch's research and edits.
The judge cited nine examples of "non-existent quotes and miscited propositions" that appeared to come from federal appellate rulings and a Congressional committee report. She also said their brief had six errors that may have been less egregious, as well as other "miscellaneous problems."
Rasch's process "included the use of Westlaw, Westlaw's AI features, Google, Google Scholar, as well as the 'deep research' feature of the Pro version of ChatGPT version 4.5," the brief said. The lawyers said Rasch used a feature on the legal research platform Westlaw called Quick Check to vet the brief, but didn't do so again after accidentally adding unvetted sections from previous drafts.
Maddux, the lawyers added, was busy with another case.
Maddux, Rasch, and Burke didn't immediately respond to requests for comment. Neither OpenAI, the company behind ChatGPT, or Thomson Reuters, which makes Westlaw, responded to requests for comment.
The proliferation of AI and the high cost of legal research has led to a number of attorneys being called to the mat by judges over errors in their legal arguments, often a result of generative AI systems' tendency to "hallucinate."
Often, the mistakes are made by solo practitioners or lawyers from small firms, though big firms have also been found using AI. A Latham & Watkins attorney said the AI system Claude was to blame for giving the wrong name and authors for an article cited in an expert's report, though the content was otherwise correct. Last week, attorneys from the firms K&L Gates and Ellis George were told to pay $31,000 after their submissions were found to contain made-up citations.
Burke, a former Deadspin editor now working as a media consultant, faces charges of hacking into a streaming system used by broadcasters. The case has attracted attention from press freedom advocates, with his lawyers arguing Burke committed no crime since the URLs he visited to download clips of Fox News footage were public.
The footage, which included antisemitic remarks by the rapper Ye and behind-the-scenes comments by Tucker Carlson about sex, his "postmenopausal" viewers, and issues with the Fox Nation streaming service, was never aired on the network. When the clips appeared online in 2022 and 2023, it aroused suspicions that a Fox employee had leaked them. In 2023, however, federal investigators zeroed in on Burke, who was indicted last year.
