Minnesota: AI Ethics Guidance for Law Firms

Bottom line for a 5-50 attorney Minnesota firm: No numbered LPRB opinion governs AI use. The most specific guidance from the discipline office is then-Director Susan Humiston’s September 2024 column summarizing ABA Formal Opinion 512, which agrees that using many generative AI tools “will generally require a client’s informed consent” and that “Generic, boiler-plate language in engagement agreements is an insufficient way to obtain informed consent.” Minnesota court rules do not require AI disclosure in filings, but in Kohls v. Ellison (D. Minn. 2025) the court excluded an expert declaration with AI-generated citations and suggested that Rule 11 may now require attorneys to ask their witnesses about AI use.

Start with: a written AI use policy, a citation verification checklist for court filings, and AI vendor due diligence records.

OLPR guidance: “Ethics Guidance for Generative AI Use”

Citation: Susan M. Humiston, Ethics guidance for generative AI use, Bench and Bar of Minnesota (September 2024), also posted on the OLPR website. Written by the Director of the Office of Lawyers Professional Responsibility (OLPR). Ms. Humiston died on August 4, 2026; the Minnesota Supreme Court appointed Binh T. Tuong Acting Director on August 6, 2026.

Status: Informal guidance in a bar journal column. It is not a Lawyers Professional Responsibility Board opinion and is not binding. The column “presents a high-level summary” of ABA Formal Opinion 512, so most of its statements describe what the ABA opinion says, sometimes with the Director’s agreement.

What it says

• Competence. The ABA opinion “does say we must have a ‘reasonable understanding of the capabilities and limitations’” of a tool before using it, and “this is not a one-and-done task.”
• Informed consent. “The opinion posits, and I tend to agree, that the use of many generative AI tools (beyond simple idea-generation tools) will generally require a client’s informed consent.” “Generic, boiler-plate language in engagement agreements is an insufficient way to obtain informed consent.”
• Communication. Lawyers sometimes “must communicate your use unprompted even when informed consent is not required,” for example when using generative AI to evaluate and advise on jury selection.
• Candor. “You should review all output that is going to be incorporated into work product presented to a tribunal for accuracy, just as you would any other sources you cite.”
• Supervision. “Whether generative AI use is permitted in your workplace should be the subject of a firm or office policy. Training should be provided if generative AI tools are used,” and lawyers should understand whether outside vendors or service providers are using generative AI.
• Fees. If clients pay for particular tools, “that must be explained at the beginning of the representation, preferably in writing,” and hourly billing “should reflect that efficiency, as you can never bill for more time than actually spent.”

Notable gaps

The column does not address disclosure of AI use to courts.

As of 2026-09-30, the Lawyers Professional Responsibility Board’s opinion index contains no opinion on AI.

MSBA AI Working Group Report

The MSBA Assembly adopted the AI Working Group’s report, Implications of Large Language Models on the Unauthorized Practice of Law and Access to Justice, on June 20, 2024 (announcement). The report is bar policy guidance, not a formal opinion or rule. In the Working Group’s estimation, the Minnesota Rules of Professional Conduct “already offer good guidance regarding LLMs,” and it observes that the line between “legal information” and “legal advice” is “largely indistinguishable” for today’s LLMs. It proposed a regulatory sandbox for access-to-justice tools in areas of acute need such as housing and immigration, and the announcement said a standing AI committee “will be formed in the near future.”

The MSBA’s AI Committee later sent a sandbox recommendation to the Minnesota Supreme Court (approved August 27, 2025). In a May 2026 Bench and Bar column, the committee’s chair wrote that the committee “continues awaiting the Supreme Court’s response to our recommendation regarding the AI Sandbox.”

What federal courts in Minnesota require for AI use in filings

On October 1, 2025, the U.S. District Court for the District of Minnesota adopted a Gen-AI FAQ on the recommendation of its Federal Practice Committee; the court’s public notice calls it “an informational resource.” The FAQ states that the court “does not currently require disclosure of AI use in court filings,” and reminds litigants that under Fed. R. Civ. P. 11 and 26(g) “you are ultimately responsible for the accuracy and sufficiency of the resulting work product.” As of September 2026, none of the D. Minn. judges’ posted practice pointers addresses AI, so any AI directive would come from a case-specific order.

What Minnesota state courts require for AI use in filings

The Minnesota Judicial Branch’s AI FAQ states: “Minnesota court rules do not currently require disclosure of AI use in court filings. However, it may be a requirement in your individual case if ordered by a judicial officer.” It adds that “you are responsible for the accuracy of anything you file in court, even if you got it from an AI tool.” None of the Minnesota Supreme Court’s rules orders issued through August 2026 addresses AI.

The Minnesota Rules of Professional Conduct govern, with Rule 1.1 Comment [8] (keeping abreast of “the benefits and risks associated with relevant technology”), Rule 1.6 Comment [17] (safeguarding client information), and Rule 5.3 (nonlawyer assistance) most relevant. The technology comment accompanied rule amendments effective April 1, 2015; Minnesota’s comments “are included for convenience and do not reflect court approval or adoption.”

State courts have addressed AI-fabricated citations case by case, including Hennepin County district court orders in Brooks v. Patel (2025) and Osborne Properties v. Mohamed (2026), and the Court of Appeals in Asiedu v. Eastlund (2026).

Kohls v. Ellison (D. Minn. 2025)

Kohls v. Ellison, No. 24-cv-3754 (LMP/DLM), ECF No. 46 (D. Minn. Jan. 10, 2025) (Provinzino, J.). The Minnesota Attorney General’s office submitted an expert declaration by Professor Hancock that contained citations to nonexistent articles; Professor Hancock acknowledged using GPT-4o. The court excluded the declaration and reminded counsel that Fed. R. Civ. P. 11 “imposes a ‘personal, nondelegable responsibility’ to ‘validate the truth and legal reasonableness of the papers filed.’” It suggested that an “inquiry reasonable under the circumstances” under Rule 11 “may now require attorneys to ask their witnesses whether they have used AI in drafting their declarations and what they have done to verify any AI-generated content.” No monetary sanctions were imposed.

The court denied the preliminary injunction in a separate order, which the Eighth Circuit affirmed (No. 25-1300, Feb. 9, 2026). That appeal concerned the injunction request, not the order excluding the Hancock declaration.

What AI-related rules are pending in Minnesota?

No AI rule proposal is pending before the Minnesota Supreme Court as of 2026-09-30. The one pending item located is the MSBA AI Committee’s sandbox recommendation, described above.

S.F. 1886 (94th Legislature) would have made it an unfair or deceptive trade practice to fail to disclose that a person is interacting with AI, with a civil penalty “not to exceed $5,000,000” and statutory damages up to $1,000. It was not enacted: its last action was a March 18, 2026 committee report re-referring it to the Commerce and Consumer Protection Committee.

How do malpractice carriers in Minnesota treat AI use?

Minnesota Lawyers Mutual published AI ethics guidance in June 2026 that includes a sample engagement-letter AI clause and states that “There is no blanket rule requiring lawyers to notify clients every time AI is used to assist with legal work.” It is risk-management guidance, not an underwriting position. No filed AI exclusion on an admitted-market LPL form has been located. Failure to verify AI-generated content remains the clearest route to exposure, as Kohls illustrates in the federal context.

What does my Minnesota malpractice carrier ask about AI at renewal?

No Minnesota malpractice carrier has published AI-specific application items in the materials reviewed for this entry. Firms that want to be ready for questions may wish to have on hand a written AI use policy, AI vendor due diligence records, an engagement letter AI clause that is matter-specific rather than boilerplate, citation verification records for court filings, and, in matters with retained experts, a record of asking experts about AI use. Confirm application items directly with the broker.

What documentation should a Minnesota firm keep on file?

Month one (foundational)

1. (Owner: managing partner + firm administrator) Written AI use policy covering approved tools, prohibited tools, what client data may be input, required verification steps, and internal violation reporting. The OLPR column says AI use “should be the subject of a firm or office policy.”
2. (Owner: litigation lead) Citation verification checklist for AI-assisted court filings, including an inquiry to retained experts about AI use and their verification procedures, as Kohls v. Ellison suggests.
3. (Owner: managing partner + billing partner) Engagement letter AI clause that is matter-specific and plain-language, since the OLPR column describes boilerplate as “an insufficient way to obtain informed consent.”

Months two and three (operational documentation)

1. (Owner: firm administrator + outside IT) AI vendor due diligence records for each tool: terms of service, data retention and training-data practices, and a documented conclusion on suitability for client-confidential information. This supports the reasonable safeguards described in Rule 1.6 Comment [17] and the supervision duties in Rule 5.3.
2. (Owner: firm administrator) Attorney and staff training records showing competence in AI capabilities, hallucination risk, and firm policy.
3. (Owner: managing partner) Supervision protocol covering associate, staff, and outside-vendor AI use under Rule 5.3, including whether outside vendors are themselves using generative AI in client-matter workflows.

Months four to six (per-matter discipline)

These are recurring practices, not one-time projects: each new matter exercises them.

1. (Owner: matter lead attorney) Per-matter AI use notes and client consent records capturing which tools were used, what data was input, and that informed consent was obtained where required.
2. (Owner: matter lead attorney) Judge-specific AI order check at the opening of each matter, since any AI directive in Minnesota courts would come from a case-specific order.
3. (Owner: billing partner) Billing transparency records documenting how AI tool costs were explained to the client and that time billed reflects time actually spent under Rule 1.5.

We are building practitioner resources for firms working through this list. The newsletter covers new resources as they ship.