USPTO Reprimands Patent Attorney Over AI-Generated Fake Citations
USPTO publicly reprimanded a patent attorney for submitting AI-generated, unverified citations.
Why it matters: This is the first reported enforcement action addressing AI use and due diligence failures in patent prosecution, underscoring the ethical and regulatory challenges patent lawyers face with AI tools.
- USPTO disciplined a patent attorney in September 2026 for using generative AI to create a response including fake legal citations.
- The attorney failed to verify AI-generated case citations, which did not exist in legal databases.
- Discipline included public reprimand and mandatory ethics-focused continuing legal education (CLE), no suspension or disbarment.
- USPTO has previously warned practitioners about risks of unverified AI output and enforces Rule 11.18(b), requiring reasonable inquiry before filing.
The United States Patent and Trademark Office (USPTO) has taken the first known enforcement action against a patent attorney for submitting unverified AI-generated material in a patent prosecution filing. The incident came to light in September 2026 when the USPTO publicly reprimanded the attorney for relying on generative AI – specifically ChatGPT – to draft a response to an Office Action that included at least three case citations that could not be verified and did not exist in applicable legal databases.
The USPTO disciplinary order summarized by the National Law Review noted that the attorney "relied on ChatGPT to generate responses and did not independently check those responses for accuracy." This failure contradicted USPTO Rule 11.18(b), which mandates that submissions be based on knowledge and information formed after reasonable inquiry.
The disciplinary measures imposed included a public reprimand and a requirement for additional continuing legal education (CLE) focused on ethics. The USPTO chose not to suspend or disbar the attorney, signaling a measured response to this emerging compliance challenge.
Patent practitioners have been repeatedly cautioned by the USPTO about the reliability issues of generative AI tools. The phenomenon known as AI "hallucination"—where AI fabricates plausible but false information—poses risks in legal filings, as seen in recent federal court disciplinary actions related to fabricated citations. Such events emphasize that attorneys must verify all AI-generated content, as their signature certifies its accuracy regardless of whether humans or AI produce the work, according to a Morgan Lewis regulatory analysis.
This case highlights the evolving regulatory landscape as AI use in intellectual property law accelerates and underscores the ethical duties patent lawyers and legal operations teams must uphold to ensure filings meet required professional standards.
By the numbers:
- 3 — Nonexistent AI-generated case citations in the attorney's Office Action response
- September 2026 — Month and year of USPTO disciplinary action
- 1 — First known enforcement action by USPTO involving AI material in patent prosecution
What's next: Watch for further USPTO guidance on AI use and verification standards in patent filings as adoption grows.