Delaware court sanctions counsel over AI-assisted brief
The Delaware Court of Chancery sanctioned counsel $2,500 under Rule 11 after a plaintiffs’ brief contained fabricated quotations and mischaracterized legal authorities. The court said generative AI is not forbidden in legal practice, but lawyers must verify the authorities and citations they submit.

Key takeaways · 4
- 01
Counsel was sanctioned $2,500 under Rule 11, payable personally to the Court of Chancery.
- 02
The court said using generative AI is not inherently problematic, but lawyers must check submitted authorities and citations.
- 03
The corrected brief removed quotation marks from nonexistent quotations but retained inaccurate descriptions of cases.
- 04
The court said Rule 11 sanctions should deter carelessness, not punish or compensate litigants.
What the court ordered
The court sanctioned counsel $2,500 under Rule 11, with payment due personally to the Court of Chancery.[1] It said Richards, Layton & Finger (RLF) demonstrated exceptional circumstances relieving the firm of joint responsibility for the sanction.[1] The court also described the purpose of Rule 11 sanctions as deterring carelessness, rather than punishing or compensating litigants.[1] For legal teams, the order distinguishes the individual lawyer’s responsibility from the firm’s potential responsibility in this case. That distinction makes the court’s findings about review and delegation especially important.
Errors in the filing
The plaintiffs filed their answering brief on January 22, 2026.[1] Five days later, defendants identified fabricated quotations, inaccurate citations and hallucinated legal propositions.[1] The court said the brief contained fabricated quotations and mischaracterized legal authorities.[1] A corrected brief removed quotation marks from quotations that did not exist, but it still inaccurately described the cases.[1] The record therefore points to more than a formatting issue: correcting the quotations did not resolve the inaccurate legal descriptions.[1] Teams should review both quoted text and how a filing characterizes precedent.
The lawyer’s review duty
Rollo used a tool to help draft and condense arguments in the answering brief.[1] The drafting process introduced fabricated quotations, flawed citations and a mischaracterized case holding.[1] Rollo delegated citation checking to a junior attorney and a paralegal, then filed without personally verifying the authorities, citations or quotations.[1] The court held that a lawyer’s duty to conduct a reasonable inquiry and be candid with the court is personal and nondelegable.[1] Delegating a check does not, under that ruling, transfer the filing lawyer’s own duty to verify what is submitted.
AI is allowed, verification is required
The court said generative AI is not forbidden or inherently problematic in legal practice.[1] It also said lawyers using GenAI must ensure the accuracy of authorities and citations they submit.[1] RLF had barred unapproved GenAI tools in March 2023, formed a working group in 2024 to evaluate tools and ethical guidance, and adopted a comprehensive policy in July 2025.[1] That policy requires an experienced lawyer to review output for accuracy and completeness, and bars using it instead of independent judgment or legal analysis.[1] More than 200 firm employees had completed its GenAI ethics training by January 2026.[1]
The ruling treats AI-assisted drafting as compatible with legal practice while keeping verification responsibility with the lawyer who submits the work. Legal teams can use the case to test whether review procedures cover citations, quotations and characterizations of precedent—not just whether an AI tool is approved.
Why it matters
AI training for your team
kju.ai is built for teams: daily sessions, team analytics, and rollout support for your organization.
How this developed
16 November 2026
Delaware court sanctions counsel over AI-assisted brief