TLDR: The Delaware Court of Chancery has issued a significant caution to litigants regarding the use of generative artificial intelligence (GenAI) in drafting court filings. This warning follows a case where a pro se litigant submitted a motion containing fabricated case citations and quotes, suspected to be ‘hallucinations’ from an AI program. The court emphasized the need for meticulous verification of AI-generated content to avoid potential sanctions.
The Delaware Court of Chancery has taken a proactive stance on the burgeoning use of generative artificial intelligence (GenAI) in legal proceedings, issuing a formal caution to litigants about its responsible application in drafting court filings. This significant guidance, delivered by Vice Chancellor Lori Will in April 2025, marks the Delaware Court of Chancery’s first direct address on the matter, stemming from the case of Daniel Jaiyong An v. Archblock, Inc. .
In the Archblock case, a pro se litigant filed a motion to compel that raised judicial eyebrows. The court observed numerous case citations that did not support the propositions for which they were cited, and, more alarmingly, direct quotes that appeared to be entirely invented by GenAI, not found in any cited cases. Independent research by the court confirmed these discrepancies, leading to the suspicion that the litigant’s GenAI tool had ‘hallucinated’—producing plausible but fabricated or inaccurate outputs .
Vice Chancellor Will’s decision, 2025 WL 1024661, at *1 (Del. Ch. Apr. 4, 2025), seized the opportunity to provide crucial guidance to all litigants within the Delaware courts. While acknowledging GenAI’s potential benefits, such as streamlining legal research, aiding in drafting, and supporting efficient document review and summarization, the court underscored the inherent risks . GenAI’s ability to enhance access to legal services, particularly for pro se litigants, was also recognized as a societal benefit .
However, the court’s primary concern revolves around the accuracy and veracity of AI-generated content. The guidance mandates that any litigant using GenAI to draft court filings must certify that they have personally reviewed the output for accuracy and verified all citations to legal authorities . This certification must be included in the filing itself. The court explicitly warned that failure to adhere to this requirement could lead to severe sanctions, including monetary penalties, the striking of filings, or even dismissal of the case . This aligns with similar orders from other courts, such as the Delaware Superior Court’s March 12, 2025, order requiring similar certification .
Also Read:
- AI’s Fictitious Legal Citations Spark Crisis in South African Courtrooms, Prompting Regulatory Action
- Intellectual Property Experts Debate Generative AI’s Fair Use Implications at 9th Circuit Conference
This ruling highlights a critical challenge in the legal profession’s adoption of advanced technology: balancing efficiency gains with the fundamental duty of accuracy and integrity in court submissions. The Delaware Court of Chancery’s caution serves as a clear directive for legal professionals and self-represented litigants alike to exercise extreme diligence when incorporating generative AI into their legal work, ensuring that the pursuit of efficiency does not compromise the foundational principles of justice and truth.
