What ethical duties must a Missouri law firm address in a policy for using generative AI?

Plain-English summary

The opinion treats generative AI like other new technology that requires an ethical assessment before law-firm implementation. Its list is expressly nonexclusive. Rule 4-1.1 requires the lawyer to learn which tools are appropriate for legal practice and to understand their risks and benefits.

Confidentiality requires assessment of a platform's terms, security, use of submitted information, and data sources. Under Rule 4-1.6(c), the firm must make reasonable efforts against unauthorized access and inadvertent or unauthorized disclosure, including disclosure by supervised people. The transmission precautions in Comment [16] also apply when use of a platform sends client information.

The lawyer must obey court orders and tribunal rules under Rule 4-3.4(c), preserve independent professional judgment under Rule 4-5.4, and verify the accuracy and content of AI-assisted work under Rule 4-5.3. Managers and supervisors must address lawyer and nonlawyer compliance under Rules 4-5.1 and 4-5.3, including training and reasonable remedial action when consequences can still be avoided or mitigated.

The policy must also consider how AI affects fee reasonableness under Rule 4-1.5(a). Because generative-AI tools are not always accurate, Rule 4-3.3(a)'s prohibition on false statements to a tribunal makes competence, supervision, review, and correction of output part of the candor analysis.

In practice

Under this opinion, adopting generative AI is not a one-rule question. The firm must evaluate the tool, its handling of client data, the people who use and supervise it, the accuracy of its output, the requirements of any tribunal, the lawyer's independent judgment, and the resulting fee.

Common questions

Q: Does competence require a lawyer to understand the AI tool?

A: Yes. The opinion requires education and training sufficient to identify appropriate uses and understand the platform's risks and benefits.

Q: What confidentiality features must the lawyer assess?

A: The platform's terms, information security, use of submitted data, sources used to generate responses, and precautions against unauthorized access or disclosure.

Q: May a lawyer rely solely on generated content?

A: No. Rule 4-5.4 requires independent professional judgment, and Rule 4-5.3 requires verification of AI-assisted work's accuracy and content.

Q: What are managers and supervisors responsible for?

A: Rules 4-5.1 and 4-5.3 require measures supporting compliance by lawyers and nonlawyers, appropriate training, and remedial action in the circumstances described by those rules.

Q: Can AI use affect the lawyer's fee?

A: Yes. The opinion directs the firm to consider fee reasonableness under Rule 4-1.5(a).

Q: How does candor to the tribunal apply?

A: Rule 4-3.3(a) prohibits knowingly making or failing to correct a false statement of fact or law, so inaccurate generated content must be caught through competence and supervision.

Background and rules framework

The opinion organizes its initial guidance around Rules 4-1.1, 4-1.6, 4-3.4, 4-5.4, 4-5.3, 4-5.1, 4-1.5, and 4-3.3. It also points to earlier Missouri informal opinions on third-party payments, metadata, device theft, crowdfunding, cloud computing, email, virtual practice, vendor relationships, misdirected email, and false information in litigation.

Citations and references

Rules of Professional Conduct:

• Missouri Rule 4-1.1 and Comment [6] (competence and technology)
• Missouri Rule 4-1.6(a) and (c), Comments [15] and [16] (confidentiality and safeguards)
• Missouri Rules 4-3.3(a) and 4-3.4(c) (candor and tribunal rules)
• Missouri Rules 4-5.1, 4-5.3, and 4-5.4 (supervision and professional independence)
• Missouri Rule 4-1.5(a) (reasonable fees)

Other opinions cited:

• Missouri Informal Opinions 2023-09, 2021-13, 2020-26, 2018-10, 2018-09, and 990007: earlier technology issues
• Missouri Informal Opinions 2021-12, 2021-03, 2018-04, 2022-07, and 2017-02: virtual practice, vendors, cloud computing, and supervision
• Missouri Informal Opinions 2020-25 and 2020-24: false information in litigation

See also

• Texas Ethics Op. 705: Generative AI in Law Practice
• North Carolina 2024 Formal Ethics Op. 1: AI in Law Practice
• Florida Ethics Op. 24-1: Generative AI in Law Practice

Source

• Landing page: Missouri Informal Opinion 2024-11

Original opinion text

Reproduced from the official source for research purposes. The linked source is authoritative.

Question: Lawyer would like to use generative artificial intelligence (AI) platforms and services in Lawyer’s practice and asks for guidance regarding whether Lawyer may ethically use this emerging technology. What ethical issues should Lawyer consider in developing a policy to use this technology in Lawyer’s practice within Law Firm?

Answer: Various forms of artificial intelligence are used by lawyers every day. However, Lawyer rightly has distinguished that generative artificial intelligence, a type of AI wherein the platform being used is learning and further developing from each query or task to generate new content and produce an appropriate response in this context to assist Lawyer, requires an examination of ethical considerations just as any other new technology service or device does that is being considered for implementation into Law Firm’s use. In developing a policy to use generative artificial intelligence platforms within Law Firm, this office provides the following initial guidance as an Informal Opinion on the subject of generative artificial intelligence. The Informal Opinion is not intended to be an exclusive list of ethical considerations, as all of the Missouri Rules of Professional Conduct pursuant to Rule should be considered, but it is intended to address key ethical considerations to the question presented. Lawyer should also analyze other legal considerations outside the scope of an informal opinion.

It is important to note that this is not the first time that an Informal Opinion has addressed technology considerations for lawyers. Lawyer may also gain guidance and understanding from reviewing other technology-related Informal Opinions including: 2023–09 (lawyers may not use third-party payment programs where advance paid legal fees or expense are not deposited directly in a client trust account); 2021–13 (metadata); 2020–26 (theft of laptop, cell phone, bar card, and credit cards; loss of client confidential information); 2018–10 (crowdfunding); 2018–09 (cloud computing); and 990007 (email).

First, Lawyer must consider the duty of competence in the appropriateness of use of generative AI. Rule 4–1.1 – Competence, states that “[a] lawyer shall provide competent representation to a client. Competent representation requires the legal knowledge, skill, thoroughness and preparation reasonably necessary for the representation.” Comment [6] to Rule 4–1.1 provides guidance that part of that duty of competence by stating that “[t]o maintain the requisite knowledge and skill, a lawyer should keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology, engage in continuing study and education, and comply with all continuing legal education requirements to which the lawyer is subject.” [emphasis added.] Lawyer should get education and training to ascertain what types of generative AI are and are not appropriate for use by Law Firm. Not all generative AI platforms and services are intended for use by lawyers, and Lawyer must understand the risks and benefits of implementing use of these technologies.

Second, Lawyer must consider confidentiality. Rule 4–1.6(a) on confidentiality generally prohibits a lawyer from revealing information related to a representation of a client unless an exception is met. That means that Lawyer needs to carefully assess any generative AI platforms or services that will be used by Law Firm to ensure confidentiality of client information is maintained. Lawyer should carefully consider such factors as the terms and conditions of using a generative AI platform or service to understand the security of the information being inputted, how that information is being used by the platform or service, and what data sources the platform or service is using to produce responses to prompts or queries. See Informal Opinion 2018–04 (cloud computing) and Informal Opinion 2021–12 (virtual practice).

Additionally, Rule 4–1.6(c) states that “[a] lawyer shall make reasonable efforts to prevent the inadvertent or unauthorized disclosure of, or unauthorized access to, information relating to the representation of the client.” Comment [15] to Rule 4–1.6 provides guidance that lawyers are required to act competently to safeguard client confidential information and creates three categories for doing so: (1) unauthorized access by third parties; (2) inadvertent or unauthorized disclosure by the lawyer or other persons who are participating in the representation of the client; and (3) inadvertent or unauthorized disclosure by those subject to the lawyer’s supervision. In describing these categories, Comment [15] to Rule 4–1.6 references Rule 4–1.1 (Competence), Rule 4–5.1 (Responsibilities of Partners, Managers, and Supervisory Lawyers), and Rule 4–5.3 (Responsibilities Regarding Nonlawyer Assistants). Comment [15] to Rule 4–1.6 also provides factors that lawyers should consider as to the reasonableness of efforts to safeguard client confidential information, and it includes a reference that state or federal data privacy laws outside the Rules of Professional Conduct may require additional safeguards over client confidential information or notification in the event of a loss of, or unauthorized access to, such information. In considering the use of a generative AI platform or service, lawyers are required to make reasonable efforts to safeguard client confidential information in accordance with Rule 4–1.6(c) and Lawyer should consider the guidance of Comment [15] as to how client confidential information will be safeguarded.

Further, Comment [16] to Rule 4–1.6 provides guidance as to reasonable precautions “[w]hen transmitting a communication that includes information relating to the representation of a client….” Lawyer should consider the guidance from Comment [16] to Rule 4–1.6 to the extent use of a generative AI platform or service may include transmission of client confidential information.

Third, to the extent court orders or court rules implicate the use of any generative AI platform or service, Lawyer should be mindful of the obligation pursuant to Rule 4–3.4(c) that prohibits knowingly disobeying an obligation under the rules of a tribunal.

Fourth, Lawyer and Law Firm must protect and maintain professional independence and independent professional judgment as required by Rule 4–5.4 and not rely solely on content created by a generative AI platform or service.

Fifth, if Lawyer or members of Law Firm use content produced with the assistance of a generative AI platform or service, just as any other time a lawyer is being assisted by a nonlawyer, there is a professional responsibility to verify the accuracy and content of the product in accordance with Rule 4–5.3. Professional responsibilities regarding nonlawyers outside the firm, including service providers, are addressed by guidance in Comments [3] and [4] to Rule 4–5.3. See Informal Opinion 2021–12 (virtual practice); Informal Opinion 2021–03 (contract with vendor for disposal of client files); and Informal Opinion 2018–04 (cloud computing). Per Rule 4–5.3(c), if Lawyer has managerial authority or supervisory authority, Lawyer is ethically responsible for conduct that would be a violation of the Rules of Professional Conduct if engaged in by Lawyer if Lawyer orders or with specific knowledge of the conduct ratifies it, or knows of the conduct at a time when its consequences could have been avoided or mitigated, but Lawyer failed to take reasonable remedial measures. See Informal Opinion 2022–07 (email sent to incorrect address); Informal Opinion 2021–12 (virtual practice); and Informal Opinion 2017–02 (lawyer’s responsibilities when a nonlawyer assistant breaches confidentiality).

Sixth, in developing this generative AI use policy, Lawyer and Law Firm should also consider supervisory responsibilities in relation to Rule 4–5.1, which requires managers and supervisors to ensure that other lawyers in the firm conform to the Rules of Professional Conduct. Per Rule 4–5.1(c), a lawyer with such responsibility within Law Firm is responsible for the conduct of another lawyer’s violation of the Rules of Professional Conduct if the lawyer orders, or with specific knowledge of the conduct ratifies it or otherwise fails to take reasonable remedial action at a time when the lawyer knows of the conduct and consequences can be avoided or mitigated. See Informal Opinion 2021–12 (virtual practice). An ethical framework for the use of generative AI, if Lawyer and Law Firm decide to use it, should be developed, and, just as with any other resource or tool used in the practice of law, appropriate training should be provided to educate lawyers and nonlawyers. See Rule 4–5.1 and 4–5.3.

Seventh, Lawyer and Law Firm should consider how use of generative AI may impact the reasonableness of fees pursuant to Rule 4–1.5(a).

Finally, use of a product of generative AI can also implicate Rule 4–3.3 – Candor Toward the Tribunal. Rule 4–3.3(a) states that “[a] lawyer shall not knowingly: (1) make a false statement of fact or law to a tribunal or fail to correct a false statement of material fact or law previously made to the tribunal by the lawyer.” At this point, generative AI tools are not always accurate, thereby requiring the careful attention to competence and supervision as outlined above to avoid any false statement of material fact or law to a tribunal. See Informal Opinion 2020–25 (remedial measures when lawyer learns of false information filed in connection with dissolution matter) and Informal Opinion 2020–24 (false testimony of client at deposition).