The Rules Now Apply to Everyone
Horst Legal Counsel | March 2026
If you have been following California’s growing body of appellate authority on AI-generated legal citations, you have watched the courts build a wall, one published opinion at a time.
First came the warnings to attorneys.
Then came the sanctions.
Then, in Torres Campos v. Munoz, the consequences reached the bench itself when fabricated case law made it into a signed court order.
Now the First District Court of Appeal has laid another brick. In Sheerer v. Panas (2026) ___ Cal.App.5th ___ (A171804), the court extended the prohibition against unverified AI-generated citations to self-represented litigants, and it made clear that the available penalties go well beyond striking a deficient brief.
The message is no longer limited to the bar. It applies to everyone who files a document in a California court.
What happened
The underlying dispute was a family law matter involving child support. Anna Sheerer appealed a trial court order that failed to account for her ex-husband Thomas Panas’s bonus and restricted stock unit compensation when calculating support. The Department of Child Support Services agreed with Sheerer that the trial court had erred. On the merits, the Court of Appeal reversed. That part of the opinion is unpublished.
What the court chose to publish was something else entirely.
In his respondent’s brief, Panas:
Cited at least two cases that do not exist.
Fabricated quotations that appear nowhere in the real cases he did cite.
Referenced hearings in the proceedings below that the appellate court could not locate in the record.
When Sheerer moved to strike the brief, Panas filed a declaration explaining that his “error” resulted from using a generative AI tool without verifying the output. The court was not persuaded by his framing.
The court’s reasoning
The First District began where you would expect: with the principle that self-represented litigants are held to the same standards as attorneys. Citing Kobayashi v. Superior Court and Keitel v. Heubel, the court rejected the notion that pro se status provides any shelter from the obligations that attach to every filing.
As our Supreme Court put it more than three decades ago, a doctrine permitting exceptional treatment for self-represented parties “would lead to a quagmire in the trial courts, and would be unfair to the other parties to litigation.”
The court then situated Panas’s conduct within the line of authority developing since generative AI tools became widely available:
Noland v. Land of the Free, L.P.: The Second District decision warning attorneys about unverified AI citations.
People v. Alvarez: Holding that honesty in court filings is paramount.
The opinion includes a passage that captures the court’s concern precisely: The rule of law depends on courts impartially applying law to facts. That foundation is threatened when litigants falsify the truth, “especially when done by using technological innovations that can spin persuasive webs of untruths and invent legal principles or authorities.”
Consequences and Penalties
Having disregarded Panas’s brief in its entirety, the court turned to consequences. It noted that under California Rules of Court, rule 8.276(a)(4), the court may impose monetary sanctions on its own motion for unreasonable violations. While the court declined to impose sanctions here due to the interests of the children involved, it published the warning as a deterrent.
Why this matters beyond family court
Sheerer v. Panas is a narrow case on its facts, but the published portion carries broad implications. Three things stand out:
Universal Obligations: The court closed the gap. Sheerer makes explicit that the same obligations apply to every litigant, represented or not.
Escalating Consequences: The court has put all parties on notice that monetary sanctions are available for AI-related citation failures. Future litigants should not expect the same restraint shown to Panas.
No “Ignorance” Defense: The court’s response to Panas’s claim that he didn’t “knowingly” submit false info was direct: his lack of knowledge resulted from a failure to verify. The duty is to confirm every citation exists and says what you claim it says.
What businesses and individuals should take away
If your company uses AI tools in any part of a legal workflow—whether for research, drafting, or document review—Sheerer v. Panas adds to the growing authority that verification is not optional.
Key Takeaway: The obligation runs to the person whose name is on the filing, regardless of whether that person is an attorney, a corporate representative, or a self-represented individual.
For Litigants: If you suspect an opposing party is submitting AI-generated work product without verification, you now have clear authority to seek relief.
For Self-Represented Parties: Understand that the tools you use do not change the standard you are held to. The consequences are getting sharper.
Bottom Line
Sheerer v. Panas confirms what the California appellate courts have been building toward for the past two years: The duty to verify every citation in every court filing is universal. Horst Legal Counsel tracks AI-related litigation developments as part of its commitment to responsible AI integration in legal practice. For questions about how these decisions affect your business or your case, contact Jason Horst at jason@horstcounsel.com.