Misuse of AI in Legal Filings: A Growing Concern
Recent developments highlight alarming instances of artificial intelligence being misapplied within the legal field. This article explores a notable case in Oregon, shedding light on the implications of AI misuse in legal documentation.
AI tools like Claude, Chat GPT, and others are being misused and creating fictitious cases and quotes in legal filings in Oregon and the rest of the U.S. Photo taken April 28, 2026.
Allison Frost / OPB
A federal judge recently imposed a fine of $110,000 on two attorneys in Oregon due to the submission of documents filled with cases and citations generated by artificial intelligence.
This fine is significant, yet it highlights a troubling trend rather than an isolated incident.
Ankur Doshi, the general counsel for the Oregon State Bar, noted that there have been about five documented court filings in Oregon that included what is referred to as AI “hallucinations.” Across the nation, this figure rises to approximately 900 instances.
An AI hallucination occurs when a generative AI model, such as Claude or ChatGPT, produces incorrect or misleading information, including entirely fabricated content.
FILE – Undated headshot of General Counsel for the Oregon State Bar Ankur Doshi.
Courtesy of the Oregon State Bar
Doshi discussed these troubling cases on OPB’s “Think Out Loud”, examining why certain lawyers persist in using AI despite its inherent risks and how this trend is reshaping the legal landscape.
He highlighted that AI misuse isn’t limited to lawyers; it also includes non-lawyers creating their own legal documents.
“Given the accessibility of generative AI tools to the public,” he stated, “we are witnessing individuals representing themselves in court, drafting their own pleadings with AI.”
Self-represented individuals could incur court sanctions, such as fines, for improper AI usage.
The degree of discipline imposed by the state bar on attorneys varies based on what Doshi termed “aggravating” and “mitigating” circumstances. These factors relate to the transparency of AI use and the manner in which issues are rectified.
If a lawyer discloses their AI usage to the court, they might face only minimal repercussions; however, those who attempt to conceal it risk penalties similar to those faced by the two Oregon attorneys, which can include fines, suspension, or disbarment.
According to an Oregon State Bar official opinion on AI tools, lawyers must maintain competence with any tools they utilize in client representation, acknowledging that AI may generate erroneous information.
Doshi observed that many attorneys continue to embrace AI, primarily due to its capacity to enhance efficiency.
“AI can notably streamline processes for attorneys who integrate it effectively,” he noted. “It offers significant capabilities for reviewing and drafting legal work.”
However, Doshi emphasized the necessity for lawyers to deepen their understanding of the AI tools at their disposal, underscoring the importance of human oversight to prevent inaccuracies.
“To utilize AI effectively—avoiding fabrications—a human element to review the work is essential,” he explained.
“Many attorneys currently lack a comprehensive understanding of how AI functions.”
Errors stemming from AI could lead to increased costs and delays for the courts, particularly when opposing counsel has to undertake additional work to scrutinize potentially fabricated cases. Furthermore, such errors could ultimately undermine the integrity of the legal system.
“It’s not just fabricated cases,” Doshi noted. “We’re seeing fabricated legal arguments as well. This undermines the very foundation of our precedent-based legal system.”
Ankur Doshi, general counsel of the Oregon State Bar, spoke to “Think Out Loud” host Dave Miller. Click play to listen to the full conversation:
Listen to audio from OPB journalists
