AI Insights: USPTO Disciplinary Action Highlights AI Risks

The use of artificial intelligence does not diminish an attorney’s professional responsibility to ensure every court filing is accurate. In a recent contribution to IPO Daily News’ “AI Insights” series, Woods Rogers Principal and Tysons Office Managing Partner Neal Seth looks at a disciplinary action involving a registered patent attorney whose federal court filing included inaccurate citations generated with the assistance of AI. 

In the case, the attorney used one generative AI tool to help draft proposed claim constructions and a second tool to review the work. Nevertheless, the filing cited nonexistent portions of the patent specification and prosecution history, and inaccurately attributed other information.

The USPTO’s Office of Enrollment and Discipline concluded that the attorney violated professional conduct rules governing competence, diligence, misrepresentation, and conduct prejudicial to the administration of justice. Although the attorney accepted responsibility and moved quickly to correct the errors, the resulting settlement imposed a public reprimand.

Neal explains that the decision expands the conversation beyond fabricated case citations. It demonstrates that attorneys must also independently verify AI-generated references to a patent’s intrinsic record and cannot rely on one AI tool to validate another. The matter further underscores OED’s authority to discipline patent practitioners for conduct outside filings made directly with the USPTO.

IPO members may access the article here.