Eminent Domain Report

We’ve come a long way from the viral “I’m not a cat” court appearance video.  Now, beyond manipulating video appearances with filters, technology allows for the creation of entire legal documents using generative AI.  However, as recent case law highlights, beyond the whimsy generated by the cat video, there are real risks associated with the uninformed or undisciplined use of these tools. 

In Quinteros v. Harbor Distributing (Cal. Ct. App., 1st. App. Dist., June 11, 2026), the First District affirmed sanctions against attorneys who filed a brief with fabricated case citations and quotations, many of which bore the hallmarks of generative AI output.

The trial court described the filing as containing “nonexistent citations” and “fictitious quotations,” ultimately imposing sanctions under California Code of Civil Procedure § 128.7.  The Court of Appeal upheld those sanctions, emphasizing that the attorneys failed to independently verify authorities before filing.

Similarly, in Lnu v. Blanche (9th Cir. June 3, 2026), the Ninth Circuit imposed sanctions, including suspension from practice, on attorneys whose filings contained AI-generated “hallucinations,” including nonexistent cases and misattributed quotations.

The court articulated a key principle:

…the rules are not violated at the point of research and drafting, but at the point of signing and filing.

As expressed, ethical violations do not occur from the use of AI alone, but from the signing and filing of documents containing false or unsupported statements.  The attorneys’ failure was compounded by lack of candor when confronted with the inaccuracies.

The potential consequences of AI misuse in litigation is not limited to attorneys.  Rather, it extends to individuals who may seek to lean on AI to proceed on a self-represented basis or to supplement advice of counsel. 

One such risk relates to a waiver of privilege over confidential information.  California law protects attorney-client communications and attorney work product, which privileges may be waived if those materials or communications are added to an AI database.  The case law in this area remains developing. 

For example, in United States v. Heppner, the United States District Court, for the Southern District of New York, held that AI-generated materials created independently by a litigant were not protected by attorney-client privilege or work product doctrine, partly because AI platforms may retain and disclose user inputs.  

By contrast, the United States District Court for the Eastern District of Michigan, in Warner v. Gilbarco, found that a self-represented litigant’s AI-generated materials could retain work product protection, reasoning that AI functions as a tool rather than a third party.

The lesson is an ongoing one, with developing jurisprudence, that requires caution in the strategic use of AI-tools in litigation to avoid the pitfalls of sanctions and potential waivers of privileges that may compromise ones position in litigation.