Can a West Virginia lawyer use artificial intelligence, including generative AI, in a law practice?

Plain-English summary

The Lawyer Disciplinary Board opens by defining AI through federal sources (15 U.S. Code § 9401 and the notes to 10 U.S. Code § 2358) and distinguishing non-generative AI (classification, prediction, decision-making that does not create new content) from generative AI (which produces text, images, or other data in response to prompts). It identifies the main areas where AI is used in law practice: technology-assisted review of documents, automated legal research, contract and document analysis, generative drafting of memoranda and briefs through large language models, and proofreading and document organization.

The Board concludes that there is nothing in the Rules of Professional Conduct that per se prohibits a lawyer from using AI as a tool. Lawyers already use non-generative AI, often without realizing it, in legal-research tools and word-processing spell- and grammar-checkers. What the opinion requires is that the lawyer's existing ethical duties be carried through to the use of AI. The opinion walks those duties one by one: competence (Rule 1.1), diligence (Rule 1.3), communication (Rule 1.4), confidentiality (Rule 1.6), supervision (Rules 5.1 and 5.3), and candor to the tribunal (Rule 3.3).

Two themes run through the analysis. First, a lawyer must understand the technology well enough to use it competently and must supervise its output rather than accept it on its face, because generative AI can "hallucinate" and present false information as true. Second, the lawyer remains the responsible decision-maker: AI is, at best, a secondary source and a supplement to the lawyer's own reasoning, never a replacement for it. The opinion also flags that some courts have adopted standing orders requiring disclosure or human verification of generative-AI use, and that lawyers should check for those jurisdiction-specific rules.

In practice

Under this opinion, a West Virginia lawyer's use of AI tools, including generative AI, is permitted but is governed by the lawyer's existing duties. The opinion holds that competence (Rule 1.1, read with the 2015 Comment 8 on keeping abreast of the benefits and risks of relevant technology) requires a basic understanding of the AI programs a lawyer uses, and that a refusal to use technology that makes legal work more accurate and efficient may itself be a failure of competent representation. Diligence (Rule 1.3) requires human input, human oversight, and human intervention to correct mistakes.

The opinion holds that, under Rule 1.4, a lawyer should discuss the decision to use generative AI with the client, obtain informed consent confirmed in writing before delegating tasks to a generative-AI process, and explain the tool's risks and limitations. Under Rule 1.6, the opinion holds that a lawyer should review a generative-AI provider's terms of use and privacy policies, inquire how client information will be stored and who can access it, and not use generative AI in a representation unless confident the client's confidential information will be secure. Under Rules 5.1 and 5.3, the lawyer is obligated to supervise the AI's work, and under Rule 3.3, the use of AI outputs that contain misrepresentations or fake citations implicates the duty of candor to the tribunal and to opposing parties and counsel.

Common questions

Q: Does West Virginia prohibit lawyers from using AI or generative AI?

A: No. The opinion concludes there is nothing in the Rules of Professional Conduct that per se prohibits a lawyer from using AI as a tool in a law practice. The lawyer's ordinary ethical duties continue to apply to that use.

Q: Do I need a client's consent to use generative AI on the client's matter?

A: The opinion concludes that, under Rule 1.4, a lawyer should consult with and obtain approval from the client before using generative AI; that consent must be informed and should be confirmed in writing, and the discussion should cover the tool's risks and limitations.

Q: Can I rely on a generative-AI tool's research or drafting without checking it?

A: No. The opinion concludes that one of the biggest risks is accepting AI work product as credible on its face, that generative AI can "hallucinate" and present false information as true, and that lawyers must make sure the work product is accurate and complete. AI should be viewed at best as a secondary source and never relied on as the primary source.

Q: What confidentiality steps does the opinion identify before using generative AI?

A: The opinion concludes that, under Rule 1.6, a lawyer should familiarize themselves with the service's terms of use and privacy policies, discuss the provider's confidentiality safeguards, ask what information will be provided, how it will be stored, what security is in place, and who will have access, and not use the tool unless confident the client's confidential information will be secure. It points to LEO 2012-01 (electronic media for file storage) as instructive.

Q: Can using AI create a problem with the court?

A: Yes. The opinion concludes that the use of generative-AI outputs containing misrepresentations of fact or law, or fake citations, implicates the lawyer's duty of candor to the tribunal under Rule 3.3, especially where a court has adopted rules or a standing order requiring disclosure of AI use or verification of AI outputs.

Background and rules framework

The opinion interprets the West Virginia Rules of Professional Conduct, which are adopted by the Supreme Court of Appeals of West Virginia and track the ABA Model Rules. It reads the duty of competence (Rule 1.1, Model Rule 1.1) together with Comment 8, adopted by the West Virginia Supreme Court in 2015, which ties competence to keeping abreast of the benefits and risks of relevant technology. It then applies the duties of diligence (Rule 1.3), communication (Rule 1.4), confidentiality (Rule 1.6), supervision of lawyers and nonlawyer assistance (Rules 5.1 and 5.3, the latter retitled in 2015 from "Nonlawyer Assistants" to "Nonlawyer Assistance"), and candor to the tribunal (Rule 3.3). It also references the duties to avoid frivolous claims (Rule 3.2), truthfulness to others (Rule 4.1), reasonable fees (Rule 1.5), prospective-client duties (Rule 1.18), communication with represented and unrepresented persons (Rules 4.2 and 4.3), and misconduct (Rule 8.4).

Citations and references

Rules of Professional Conduct:

• Model Rule 1.1 / WV RPC 1.1 (competence, including Comment 8 on technology)
• Model Rule 1.3 / WV RPC 1.3 (diligence)
• Model Rule 1.4 / WV RPC 1.4 (communication)
• Model Rule 1.5 / WV RPC 1.5 (fees)
• Model Rule 1.6 / WV RPC 1.6 (confidentiality)
• Model Rule 1.18 / WV RPC 1.18 (duties to prospective clients)
• Model Rule 3.2 / WV RPC 3.2 (expediting litigation)
• Model Rule 3.3 / WV RPC 3.3 (candor toward the tribunal)
• Model Rule 4.1 / WV RPC 4.1 (truthfulness in statements to others)
• Model Rules 4.2 and 4.3 / WV RPC 4.2, 4.3 (communication with represented and unrepresented persons)
• Model Rules 5.1 and 5.3 / WV RPC 5.1, 5.3 (supervisory responsibilities; responsibilities regarding nonlawyer assistance)
• Model Rule 8.4 / WV RPC 8.4(c), 8.4(d) (misconduct)

Statutes:

• 15 U.S. Code § 9401 (definition of "artificial intelligence")
• 10 U.S. Code § 2358 (notes defining "artificial intelligence")

Other opinions cited:

• WV Legal Ethics Opinion 2012-01: Use of Electronic Media for File Storage (cloud computing and electronic storage; cited as instructive on confidentiality and competence)

See also

• ABA Formal Op. 512: Generative AI tools
• TX Ethics Op. 705: Generative AI in the practice of law
• Alaska Ethics Op. 2025-1: Generative AI and the practice of law
• FL Bar Ethics Op. 24-1: Generative AI use by lawyers

Source

• Landing page: https://wvodc.org/Legal-Ethics-Opinion
• Original PDF: https://storage.googleapis.com/msgsndr/Rgd68xOkcVdteTsBkf6O/media/667ac9c219bb7a1f7a4df4c2.pdf

Original opinion text

Best-effort transcription from a scanned PDF. Minor errors may remain; the linked PDF is authoritative.

LEGAL ETHICS OPINION 24-01
ARTIFICIAL INTELLIGENCE

Artificial Intelligence ("AI") is an evolving technology, and as it becomes more advanced, AI's functions and its users will continue to grow, and the legal field is no exception. According to 15 U.S. Code § 9401 AI is defined as "a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments." The notes in 10 U.S. Code § 2358 define AI as:

1. Any artificial system that performs tasks under varying and unpredictable circumstances without significant human oversight, or that can learn from experience and improve performance when exposed to data sets.

2. An artificial system developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action.

3. An artificial system designed to think or act like a human, including cognitive architectures and neural networks.

4. A set of techniques, including machine learning, that is designed to approximate a cognitive task.

5. An artificial system designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.

There are essentially two models of artificial intelligence, non-generative and generative. Non-generative AI models perform computations based on input data and focus on tasks such as classification, prediction, and decision-making. This type of AI doesn't generate any new content, but instead automates tasks while learning and adapting. Examples of models using non-generative AI: tools to detect and correct grammatical errors and spelling mistakes in written text; tools to detect fraudulent transactions in financial systems; systems to identify and filter out spam emails; and recommendation systems to suggest personalized content or products based on a user's preferences and past behavior. Conversely, generative AI can generate text, images, videos, or other data using generative models, often in response to prompts. Generative AI models learn the patterns and structure of their input training data and then generate new data that has similar characteristics. Examples of models using generative AI; Chat GPT; Lex Machina; chatbots and other conversational agents that can interact with users in natural language; 3D Modeling; and video and image creation and editing. The information available to a generative AI product is confined to the dataset on which it has been trained, which may render it incomplete, out of date or biased.

Both non-generative and generative AI are being used in the practice of law. The primary areas where AI is being applied in the practice of law include the following categories: review of documents for discoverable or otherwise relevant information, generally referred to as technology-assisted review (TAR); legal research through automated searches of a universe of case law, regulations, and statutes; contract and legal document analysis; generative drafting of legal memoranda and case briefs through the use of large language models; and proofreading, error correction and document organization.[1]

There is nothing in the Rules of Professional Conduct that per se prohibits a lawyer from using AI as a tool in a law practice. However, lawyers are cautioned to check for jurisdiction specific rules or standing orders to determine whether they are prohibited from using generative AI to prepare pleadings filed in that court or otherwise are required to disclose the use of generative AI and/or to certify human verification of the pleadings filed in the court. Lawyers use non-generative AI without even realizing it, such as in legal research tools or with spelling and grammar tools associated with word processing programs. Lawyers are increasingly using forms of AI in their practices to improve the efficiency and accuracy of legal services offered to their clients. However, to remain compliant with their ethical obligations, lawyers must not only be aware of AI's ability to deliver efficient legal services to clients, but they must also take care to exercise independent judgment, communicate with clients, maintain client confidentiality, ensure fees and costs are reasonable,[2] comply with advertising regulations,[3] and supervise the work performed by AI. The Rules of Professional Conduct were drafted with the intent of being adaptable to modern times and its technologies, and AI is no different.

Duty of Competence

Rule 1.1 of the Rules of Professional Conduct provides that "[c]ompetent representation requires the legal knowledge, skill, thoroughness and preparation reasonably necessary for the representation." In 2015, our Supreme Court adopted Comment 8 to Rule 1.1 which states that "[t]o maintain the requisite knowledge and skill, lawyers must keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology ..." Reading Rule 1.1 and Comment 8 together indicates that lawyers have an ethical obligation to keep up to date on the technology used in the legal field to provide competent representation to clients. Numerous States have instituted their own guidelines for rules governing lawyer competency on the use of technology -- like our specific requirements for continuing legal education hours in legal ethics, substance abuse, or law office management. A refusal to use technology that makes legal work more accurate and efficient may be considered a refusal to provide competent legal representation to clients.

Lawyers may incorporate AI tools into their practices but must do so without compromising the competent representation of their clients. AI can be an effective tool in a lawyer's arsenal, but the unsupervised use of AI can have catastrophic results. To ensure competent representation, lawyers should have a basic understanding of the AI programs that may best benefit their clients and their practice. AI programs are still machine-based and do not have the same level of understanding and judgment as a human lawyer when it comes to interpreting the nuance of legal principles and precedent. If the lawyers choose to use AI programs, it should be to supplement, not replace their own legal reasoning. Lawyers must understand that generative AI can hallucinate.

Duty of Diligence

Rule 1.3 requires lawyers to act with reasonable diligence and promptness in representing a client. To fulfil the lawyer's duty of diligence, a lawyer's use of AI resources requires human input, human oversight, and when necessary human intervention to correct mistakes. AI resources utilized by a lawyer should be regularly monitored and assessed to minimize the risk of errors or biases that could compromise the quality of the lawyer's work product. Lawyers are responsible for their work product and the legal advice they render, and this includes taking any corrective measures as needed from the errors that result from the use of AI.

Duty to Communicate

Rule 1.4 governs a lawyer's duty to communicate with clients and requires a lawyer to "reasonably consult with the client about the means by which the client's objectives are to be accomplished." A lawyer's duty of communication under Rule 1.4 includes discussing with his or her client the decision to use generative AI in providing legal services. Disclosures may be challenging to draft and navigate if the lawyer and the client do not have the same understanding of AI programs. Specifically, a lawyer needs to consult with a client prior to delegating certain tasks to a generative AI program or process, like a lawyer's responsibilities when outsourcing legal support services. A lawyer should obtain approval from the client before using generative AI, and this consent must be informed and should be confirmed in writing. The discussion should include the risks and limitations of the generative AI tool. In certain circumstances, a lawyer's decision not to use AI also may need to be communicated to the client particularly, if using AI would benefit the client.

Duty of Confidentiality

Under Rule 1.6, lawyers owe their clients a duty of confidentiality and this duty specifically requires a lawyer affirmatively to "make reasonable efforts to prevent the inadvertent or unauthorized disclosure of, or unauthorized access to, information relating to the representation of a client." Lawyers using generative AI services (or any services on behalf of clients) should familiarize themselves with its terms of use and privacy policies before using the service. The use of some generative AI tools may require client confidences to be "shared" with third-party vendors and this use creates a risk that confidential information may be exposed. For example, some generative AI models may store and use confidential information provided by lawyers in order to "teach" the model and/or it may be produced by the model to other parties as a result of similar searches. As a result, lawyers must take proper steps to ensure that their clients' information is safeguarded. To minimize the risks of using generative AI that utilizes client information, a lawyer should discuss the confidentiality safeguards used by the providers. A lawyer should further inquire about what type of information is going to be provided, how the information will be stored, what security measures are in place with respect to the storage of the information, and who is going to have access to the information. An existing legal ethics opinion relating to cloud computing and electronic storage disposal has addressed the duties of confidentiality and competence to technological innovations and is instructive. (See Legal Ethics Opinion 2012-01 USE OF ELECTRONIC MEDIA FOR FILE STORAGE.) Generative AI should not be used in the representation unless the lawyer is confident that the client's confidential information will be secure.

Duty to Supervise

Under Rules 5.1 and 5.3, lawyers have an ethical obligation to supervise lawyers and nonlawyers who are aiding lawyers in the provision of legal services to ensure that their conduct complies with the Rules of Professional Conduct. In 2015, our Court adopted changes to the Rules of Professional Conduct, one of which included changing the title of the Rule from "Responsibilities Regarding Nonlawyer Assistants" to "Responsibilities Regarding Nonlawyer Assistance." Rule 5.3 encompasses nonlawyers and the lawyer's attendant duty to supervise those nonlawyers --whether human or not. Under Rules 5.1 and 5.3, lawyers are obligated to supervise the work of the AI used in the provision of legal services and understand the technology well enough to ensure compliance with the lawyer's ethical duties.

One of the biggest risks associated with a lawyer's use of generative AI is overestimating the capabilities of the software or accepting the AI's work product on its face as credible. Lawyers have a duty to make sure the work product produced by AI is accurate and complete. The failure to do so can lead to violations of the lawyer's duties of competence (Rule 1.1), avoidance of frivolous claims and contentions (Rule 3.2), candor to the tribunal (Rule 3.3), and truthfulness to others (Rule 4.1), in addition to sanctions that may be imposed by a tribunal against the lawyer and the lawyer's client. Generative AI can "hallucinate"--that is, generative AI can confidently convey incorrect or misleading results and can present false information as true. Generative AI may also produce information that is subject to copyright law. The use of generative AI in any capacity in the practice of law requires due diligence by the lawyer to avoid errors and potentially disastrous results. Lawyers should view AI at best as a secondary source and should never be relied upon as the primary source itself. Lawyers still need to make final substantive decisions on the exact content of and language used after reviewing the suggestions from AI.

Duty of Candor to the Tribunal

The use of generative AI outputs that contain misrepresentations of fact or law, or that provide fake citations, implicates the lawyers' duty to be candid with the tribunal and to the opposing party and counsel. This is especially true if the tribunal has adopted rules or procedures or has issued a standing order requiring the disclosure of the use of generative AI and/or the verification or other safeguards with respect to generative AI outputs.

CONCLUSION

Technology such as AI can be used as a complement to a lawyer's work and while there is nothing inherently improper in using AI or generative AI, a lawyer's duties to their clients, the courts, and the profession under the Rules of Professional Conduct remain unaltered. While AI can streamline time-consuming and mundane tasks, lawyers must still utilize their skills in conjunction with their individual professional and moral judgment and in short, continue to be the lawyer.

APPROVED by the Lawyer Disciplinary Board on the 14th day of June, 2024, and ENTERED this 14th day of June, 2024.

Nicole A. Cofer, Chairperson
Lawyer Disciplinary Board

[1] Additionally, while the parameters of this advisory opinion are limited to the professional obligations of a lawyers' use of AI, the Board notes that the use of generative AI may offer potential solutions to reduce the justice gap. Access to justice and access to reliable legal information are closely intertwined. Accordingly, AI's utility in this regard will depend on its accuracy, its ability to preserve privacy and confidentiality, and its avoidance of biases.

[2] Generative AI programs may make a lawyer's work more efficient, but this increase in efficiency must not result in double billing for services or falsely inflated claims of time. (See Rule 1.5, 8.4(c) and 8.4(d) of the Rules of Professional Conduct.)

[3] A lawyer should be careful when using generative AI chatbot for intake purposes as the lawyer will be responsible in the event the chatbot provides misleading information to prospective clients (See Rule 1.18 of the Rules of Professional Conduct) or otherwise communicates in a manner that is violative of the Rules of Professional Conduct (See generally Rule 4.2 and Rule 4.3 of the Rules of Professional Conduct.)