Seither & Cherry Quad Cities, Inc. v. Oakland Automation, LLC and AP Electric, Inc. v. Oakland Automation, LLC (consolidated)

What sanction did the court impose?

July 28, 2025 Consolidated Opinion and Order Imposing Sanctions for Use of Artificial Intelligence (DE 92): $1,485 monetary sanction payable directly by plaintiffs' counsel (not plaintiffs) to defense counsel. The amount equals defense counsel's declared costs related to the AI-citation issue. Court framed the penalty as deterrence; noted preference for paying the directly injured party rather than the court despite Rule 11's default of paying into court. Court also encouraged but did not require plaintiffs' counsel to complete a CLE course on the use of LLMs in legal research and legal ethics.

Why does Seither & Cherry Quad Cities, Inc. v. Oakland Automation, LLC and AP Electric, Inc. v. Oakland Automation, LLC (consolidated) matter for law firms using AI?

Seither & Cherry / AP Electric v. Oakland Automation is E.D. Mich.’s representative measured AI-citation sanctions order for represented counsel, with three doctrinally distinctive features:

Rule 11 vs. inherent power separation. Judge Behm explicitly distinguished the two bases for sanctions: inherent-power sanctions require bad faith (which the court did not find), but Rule 11 sanctions can be imposed regardless of whether an error was made in good or bad faith. This is the clean version of the doctrinal point that has bedeviled many AI-citation cases. Counsel briefing AI sanctions issues in E.D. Mich. or any jurisdiction following Sixth Circuit precedent should anticipate that no-bad-faith findings do not defeat Rule 11 liability.

Real-case-with-fake-quote insidiousness. The court’s reasoning that real-case-name + fabricated-quotation citations are particularly insidious (because the real case names trigger less scrutiny than entirely fabricated cases) is a forward-looking warning. As AI tools “improve” by ensuring at least real case citations, the verification duty for quotations and explanatory parentheticals becomes more important than the existence check.

Sanction calibrated to specific defense costs. Defense counsel submitted a declaration totaling $1,485 in costs related to the AI-citation issue; the court ordered that exact amount. Behm noted his preference for paying the directly injured party rather than the court, departing from Rule 11’s default of paying into court. The remedy aligns the penalty with the actual harm.

The court accompanied the sanctions order with a notable LLM-as-word-guesser explanation, citing a Mark Riedl Medium primer on large language models and noting that LLMs “are word guessers; they are trying to predict what the next words would be if that sentence appeared on the internet. It has no way of ‘knowing’ whether that sentence it created about the case’s holding was in fact true or false.” This reasoning, accessible to non-technical readers, is useful for educating partners and clients on why AI-generated citations require independent verification regardless of how confident the output appears.

R&G’s tracker dates this entry 2025-06-05, which appears to refer to an earlier intermediate communication off the record where the court first flagged the hallucinations. The actual signed sanctions order is dated 2025-07-28 (DE 92). The Behm chambers also handles Jarrus v. Governor of Michigan, a separate pro se matter that resulted in $600 in sanctions ($300 per plaintiff), tracked separately in this database.