Case Intel:

• Indiana federal judge recommends $7,500 sanctions against employment attorney for submitting discovery brief with fabricated case citations
• Attorney blamed paralegal's rushed work but judge found "most logical explanation" was generative AI usage
• Sanctions include financial penalty, chief judge referral, and mandatory client notification within 14 days

An Indiana federal judge has recommended sanctioning an employment attorney $7,500 after discovering fabricated legal citations in a discovery brief, highlighting the persistent challenges lawyers face when incorporating artificial intelligence into their practice without proper safeguards.

U.S. Magistrate Judge Mark J. Dinsmore issued the sanctions recommendation Tuesday against attorney Tae Sture of Sture Legal Services LLC, who represents a woman suing Marion County Superior Court Juvenile Detention Center for employment discrimination. The judge found Sture's response to a motion to compel contained what he called a "rather bizarre amalgamation of several cases" and citations to nonexistent legal authority.

The sanctions stem from litigation filed by Indiana resident Yvonne L. Davis against the Marion County Superior Court Juvenile Detention Center and its chief operations officer. Davis alleges Title VII sex discrimination and First Amendment retaliation after being terminated in 2022 for alleged "untrustworthiness" and subsequently denied rehire.

When defendants filed a July 28 motion to compel discovery, Sture had just three days to respond. His July 31 brief included problematic citations that caught Judge Dinsmore's attention during review.

During a show cause hearing, Sture attributed the faulty citations to his paralegal's rushed drafting of the document under deadline pressure, denying any use of generative AI tools. However, Judge Dinsmore remained skeptical.

"The most logical explanation for the citation to non-existent authority is, of course, the use of generative AI to conduct legal research and/or draft the brief," Judge Dinsmore wrote in his recommendation.

Crucially, the judge emphasized that AI usage itself wasn't the problem. "There is nothing fundamentally improper in the use of AI tools to draft a brief," he noted. "Rather, it is counsel's abdication of his responsibility to ensure that the information he provided to the court was accurate that is the basis for the sanctions recommended."

The decision echoes a growing number of federal and state rulings over the last three years where lawyers have stumbled by submitting filings with fabricated authorities. Judges have made clear they are not anti-AI—what they oppose is sloppy lawyering that fails to validate what is presented as binding law.

Judge Dinsmore's recommended sanctions include:

• $7,500 personal financial penalty against Sture
• Referral of Sture's conduct to the court's chief judge
• Mandatory provision of sanction order copy to client Davis
• Required certification of client notification

Sture has 14 days to file objections to the recommendation.

The judge found Sture violated Indiana Rules of Professional Conduct and Federal Rule 11 by failing to conduct reasonable inquiry into the law before filing his brief, regardless of whether AI generated the problematic citations.

The defendants in the case are represented by Matthew Rayman, Caitlin Wilkinson, Jessica Reagan Gastineau, and Kiely C. Keesler of the City of Indianapolis Office of Corporation Counsel.

The case is Davis v. Marion County Superior Court Juvenile Detention Center et al., case number 1:24-cv-01918, in the U.S. District Court for the Southern District of Indiana.

Additional source material:

Across the country, judges are repeatedly forced to remind attorneys that while AI can accelerate drafting, it cannot replace due diligence. The embarrassment of having to explain phantom cases in court, combined with the financial sting of sanctions, is pushing law firms to look for safer, more reliable ways to integrate AI into their workflow.

This is where new platforms are quietly emerging. One such effort is the development of Advanced Drafting, Analysis & Methods (A.D.A.M.), a legal AI-agent currently being built with templates designed to prevent hallucinations and flag unverifiable citations before filings ever reach the court’s docket. With built-in human oversight to validate AI output, the idea is to give lawyers a protective layer that keeps them in compliance with professional standards while still benefiting from the speed of generative tools.

For those who have watched lawyers face sanctions for avoidable mistakes, the value of such safeguards is becoming increasingly clear. Attorneys who ignore the trend risk not just money but reputational harm. Those who adapt may save themselves, and their clients, from the kind of courtroom embarrassment now facing Sture.

More information about A.D.A.M.’s early-stage templates and oversight features can be found on Samuel Lopez’s Patreon page.