June 1, 2026 (in 3 days): New York: 22 NYCRR Part 161 takes effect, system-wide AI policy for all UCS courts

Rotharmel v. Dauphin County Children & Youth Services

U.S. District Court, Middle District of Pennsylvania · M.D. Pa. · Pennsylvania bar guidance

Conduct

Pro se plaintiffs filed a docket letter claiming counsel by 'an elite law firm' that was actually Callidus AI, an AI research site.

Consequence

Cautionary footnote in the R&R; warning that future misrepresentation of counsel status will not be tolerated.

Lesson

AI hallucination harm extends beyond fabricated citations to fabricated representation status. The exposure surface is wider.

Court sanction

Verified May 6, 2026

Citation
Rotharmel v. Dauphin Cnty. Children & Youth Servs., No. 4:25-CV-235 (M.D. Pa. July 21, 2025) (Bloom, M.J.)
Decided
July 21, 2025

Summary

Pro se plaintiffs filed a letter (Doc. 12) on the docket purporting to be from "an elite law firm" representing them. Chief Magistrate Judge Daryl F. Bloom, in a Report and Recommendation that recommended dismissal of the amended complaint as duplicative of an earlier-filed action, found the letter appeared to have been generated using artificial intelligence and that "Callidus AI," named in the letter as the firm representing plaintiffs, was actually an AI-generated legal research website rather than a firm of record. The court included a cautionary footnote warning that any counsel claiming representation must enter an appearance on the docket and that filings misrepresenting representation by counsel will not be tolerated.

AI tool:
Callidus AI
This case summary is informational only. Verify the underlying opinion or order against the primary source before relying on it in any filing or client matter.

What sanction did the court impose?

Report and Recommendation issued recommending dismissal on duplicative-claim grounds. Cautionary footnote on the AI-generated letter included as prospective warning. No monetary sanction imposed in this order.

Why does Rotharmel v. Dauphin County Children & Youth Services matter for law firms using AI?

Rotharmel is the case the AI-and-lawyers beat has been waiting for and largely missing. The dominant hallucination story since Mata v. Avianca has been the fabricated citation: a real attorney files a real brief that contains made-up cases. Rotharmel sits in a different category. The fabrication here is the legal representation itself. Pro se plaintiffs filed a letter on the federal docket claiming to be represented by “an elite law firm,” and the named firm turned out to be an AI legal research website, Callidus AI, that had auto-generated the letter language. The court did not need to reach the citation-fabrication question because the threshold problem was that the firm of record was not a firm.

The exposure surface this opens is broader than cite-checking captures. AI tools marketed for legal research that produce firm-letterhead-styled outputs create a misrepresentation pathway for unrepresented or marginally-represented parties who do not understand the line between a research aid and a representing entity. For the Pennsylvania bar’s purposes, the question is whether the unauthorized practice rules and the misrepresentation-of-counsel-status rules contemplate this fact pattern at all. For the small-firm administrator’s purposes, the operational implication is narrower: any inbound communication claiming counsel status should be verified against the docket and the state bar’s licensed-attorney directory before being treated as legitimate adversary correspondence. The cost of that verification is minutes; the cost of treating a Callidus-AI-style letter as real counsel correspondence is a potential ethical or competence question down the line.

Implications for your firm

Operational steps a firm reading this case may wish to consider documenting. Strategic and rule-application calls belong to your firm's attorneys.

  • Train staff to verify any inbound correspondence claiming counsel status against the docket and the state bar's licensed-attorney directory before treating it as adversary communication.
  • Document a process for handling client-side use of AI tools that auto-generate firm letterhead or attorney attribution; flag any letter purporting to be from counsel that is not entered on the docket.
  • Consider how AI legal-research products that style themselves as firm-equivalents may create unintentional misrepresentation risk for unrepresented or marginally-represented parties.

Sources

Primary sources