Technology & Courts
Judge Orders Plaintiff to Redraft Federal Lawsuit After Finding Signs of AI Use
An Arizona judge dismissed a sprawling complaint without prejudice and required a clearer amended filing, highlighting accountability rules for AI-assisted legal work.
A complaint is dismissed, not the underlying claims
A federal judge in Arizona dismissed an employment-discrimination complaint against the Agriculture Department after concluding that the 67-page filing showed hallmarks of artificial-intelligence generation and failed basic pleading standards. U.S. District Judge Krissa Lanham described the allegations as repetitive, difficult to follow and poorly connected to the legal claims. The self-represented plaintiff, Shelly George, may file an amended complaint, so the ruling does not decide whether the alleged retaliation or hostile work environment occurred.
The unusual drafting condition
Lanham directed George to personally draft the relevant allegations for the amended complaint rather than use AI to generate them. The judge did not impose a permanent ban on all future AI assistance in the case. Instead, both sides must disclose later use and confirm that outputs were verified. George told Reuters that AI had been used only in part and that she intended to refile. The dispute illustrates how courts are focusing on responsibility for the final document rather than technology alone.
Why length can become a legal problem
There is no simple page limit that automatically invalidates a federal complaint, but procedural rules require a short and plain statement connecting facts to legal claims. Excessive detail can obscure dates, actors and requested relief. Judges sometimes dismiss so-called shotgun pleadings because defendants cannot tell which allegations support which counts. AI systems may amplify the problem by producing plausible but repetitive language. A concise chronology and clearly labeled claims are usually more useful than a long narrative containing every grievance.
AI assistance is not inherently improper
Lawyers and self-represented litigants may use technology for research, organization and drafting when local rules permit. The user remains responsible for citations, factual accuracy and legal relevance. Courts across the country have sanctioned filings containing invented cases or incoherent arguments, prompting some judges to require disclosure. Others worry that categorical bans could restrict access to useful tools. The central principle is verification: a person signing a filing certifies that it has a factual and legal basis after reasonable inquiry.
Access to justice and unequal resources
People without lawyers often struggle to translate real experiences into formal pleadings. Generative tools can help organize facts or explain procedure, but they may also create confident errors and unnecessary complexity. Courts can reduce risk with plain-language forms, limited-scope legal assistance and clear instructions about permitted technology. Litigants should use official court resources and legal-aid services where available. An AI system cannot evaluate credibility, conduct discovery or replace advice tailored to a person's jurisdiction and deadlines.
Practical lessons from the ruling
Anyone using AI for a legal draft should independently read every sentence, verify every authority in an official database and remove material unrelated to the required elements. Dates, names and requested remedies should be easy to locate. Disclosure requirements must be followed exactly. The amended complaint in this case will show whether the plaintiff can present a compliant version. More broadly, appellate decisions and local rules will determine whether courts converge on a common standard or continue using different approaches to AI-assisted filings.
Reporting note: This article draws on public records and verified reporting; material claims are attributed in the text.