In a recent hearing, the Department of Justice (DOJ) mistakenly cited a non-existent case, Taylor v. Hott, to argue that a man detained by Immigration and Customs Enforcement (ICE) should not be permitted to challenge his detention. The cited case, supposedly supporting the idea that district courts cannot review immigration bond determinations by judges, turned out to be a fabrication, possibly created by generative artificial intelligence (AI).

Izzeddin Ahmed Abdulghaffar Daghra, the detainee in question, was initially set a $35,000 bond by an immigration judge, a decision the Trump administration did not honor. Instead, the government appealed, activating a 90-day automatic stay on the bond under regulation 8 C.F.R. § 1003.19(i). This maneuver allowed the government to extend Daghra’s detention past the expiration of the bond stay without acknowledging the bond order, a situation Chief Judge Hala Y. Jarbou described in an order to show cause why Daghra’s habeas petition should not be granted due to the lack of a legal basis for his continued detention.

Upon investigating the government's citation of Taylor v. Hott, Judge Jarbou discovered not only that the case did not exist at the specified reference but that the page belonged to a completely unrelated case, Atkins v. CGI Technologies & Solutions, Inc., concerning commercial arbitration. This led to the revelation that the citation might have been generated by AI without proper verification by the DOJ attorneys.

Despite the serious nature of citing fictitious legal precedents, Judge Jarbou decided against imposing sanctions. However, she highlighted the "improper and unacceptable" use of AI-generated content in legal filings and reminded the government of its duty of candor to the tribunal.

The DOJ has faced criticism for its handling of legal staff and cases during the Trump administration, including high turnover and ethical concerns. This incident adds another layer of scrutiny regarding the department's practices and the challenges of integrating AI tools in legal proceedings.

The misuse of AI in legal contexts raises significant ethical and procedural questions, especially as the technology becomes more pervasive. The legal community is still grappling with the implications of AI in practice, including the reliability of AI-generated information and the oversight necessary to prevent such errors.

This case not only underscores the ongoing issues within the DOJ but also serves as a cautionary tale about the reliance on AI without adequate human oversight in critical legal decisions. As AI continues to be integrated into various sectors, the need for rigorous checks and balances becomes increasingly apparent to prevent similar incidents in the future.