Be cautious, be curious, be vigilant, and be brave. Those were the words of the New York State Bar Association’s Task Force on Artificial Intelligence. It is good advice. GenAI is here and lawyers need to understand its risks, its benefits, and how it can (and should) be ethically integrated into their legal practice.
Despite the changing world around us, a lawyer’s fundamental ethical obligations remain the same. Attorneys are required to comply with the Rules of Professional Conduct in their state and apply those rules to GenAI. Some states have issued AI-specific guidance while others are still racing to catch up. We highlight some of the state guidance as well as the ABA Model Rules.
GenAI in ABA Model Rules
While the ABA Model Rules do not apply to attorneys directly, they form the foundation of the majority of state rules so guidance from the ABA is helpful. The relevant obligations in these rules include:
Competence (Rule 1.1)
Rule 1.1 mandates competent representation, including the legal and technological knowledge, skill, thoroughness, and preparation reasonably necessary for representation. In the context of AI, this means that attorneys must understand the capabilities and limitations of the AI tools they use, ensuring these tools enhance their ability to serve their clients effectively without compromising the quality of their work.
Confidentiality (Rule 1.6)
This rule emphasizes an attorney’s duty not to reveal information relating to the representation of a client unless the client gives informed consent. When using AI tools, attorneys should understand that GenAI programs often use or store the input data and must take appropriate steps to evaluate the tools’ policies and protect client information.
The ABA’s Comment to Rule 1.6 reinforces the obligation to safeguard information relating to a representation against unauthorized access by third parties. As technology evolves, so does the requirement for attorneys to stay informed about the security measures needed to protect sensitive client information handled by AI systems.
Supervision (Rules 5.1 and 5.3)
Attorneys have an obligation to ensure that the conduct of those they supervise complies with ethical obligations. This responsibility extends to supervising AI tools and the personnel who use them. For example, if a paralegal inputs a client’s confidential case strategy into an AI program for grammar and spell check, the supervising attorney could be held responsible for breach of client confidentiality.
State Bar Guidance on GenAI
Now, let’s take a look at some of the state-specific guidance:
California
Perhaps unsurprisingly, tech-savvy California led the charge with AI guidance for its lawyers. On November 16, 2023, the State Bar of California issued Practical Guidance on the Use of Generative AI in the Practice of Law. This guidance applies existing Rules of Professional Conduct to AI issues, including:
Confidentiality: A lawyer should ensure any generative AI tools used have adequate security and confidentiality protections, and avoid inputting client confidential information into AI systems. Lawyers must review terms of use and consult with an IT professional to ensure the security of any AI system that uses confidential client information.
Competence and Diligence: It is well-known that a lawyer’s duty of competence includes technological competence. Lawyers must understand generative AI to a reasonable degree before using it and should review AI-generated outputs for accuracy and bias. Al should never replace a lawyer’s professional judgment.
Supervision: Lawyers have a duty to train and supervise employees using AI tools and to properly supervise the work of AI systems, just as they would junior lawyers or nonlegal professionals.
Communication: The obligation to communicate always requires a lawyer to follow a client’s instructions and may require disclosure of the use of AI tools, however, the guidance does not mandate such disclosure.
Candor: A lawyer’s duty of candor to the tribunal requires that a lawyer check the accuracy of any AI generated outputs and complies with any rules regarding AI in that jurisdiction.
Discrimination and Bias: Lawyers should be aware of potential biases in AI systems, engage in continuous learning, and establish policies to identify and address them.
Compliance with the Law: Lawyers must analyze the relevant laws and regulations in each jurisdiction where they are licensed to ensure compliance when using AI.
Florida
Then came Florida with Florida Bar Ethics Opinion 24-1 in January 2024.
Confidentiality: Lawyers must understand the confidentiality issues in using GenAI and take reasonable steps to prevent inadvertent or unauthorized disclosure of client information by the AI system. It is advisable to obtain a client’s informed consent before using a third-party AI system that would involve disclosing client confidential information. Existing ethics opinions regarding cloud computing and remote support services are also instructive about this issue.
Oversight: A lawyer has an obligation to exercise appropriate supervision and control over the work of the AI system, just as they would for a non-lawyer assistant. Existing advice regarding nonlawyer assistants is useful for understanding this issue. Chatbots that use GenAI are a particular concern here and should not provide legal advice or misrepresent their limitations.
Legal Fees and Costs: Lawyer can ethically only charge a client for actual costs incurred. Increased efficiency as a result of the use of AI must not lead to inflated billing. If a lawyer intends to charge a client for the use of AI in their matter, they must inform the client about that in writing. Also, a lawyer may not charge the client for their time spent learning how to use the AI tool.
Lawyer Advertising: Lawyers must ensure that any statements about the use of AI in advertising or solicitation are not false, misleading, or deceptive. If using a chatbot or other AI tool, prospective clients must be told that they are communicating with an AI program not a lawyer or law firm employee to avoid any confusion or deception.
New Jersey
Up next is the Garden State. In January 2024, the New Jersey Supreme Court Committee on Artificial Intelligence issued its Preliminary Guidelines On The Use Of Artificial Intelligence By New Jersey Lawyers, reinforcing that a lawyer’s fundamental duties and ethical obligations remain unchanged:
Accuracy and Truthfulness: A lawyer has a duty to be accurate and truthful. AI can generate false information (“hallucinations”) therefore a lawyer has an ethical duty to check and verify all information generated by AI to ensure that it is accurate.
Honesty, Candor, and Communication: Lawyers remain responsible for the validity of legal submissions, including those generated using AI, and must not submit false, fake, or misleading content, whether generated by AI or otherwise. Although a lawyer does not have to tell a client everytime they use AI, they may have an obligation to disclose the use of AI if the client cannot make an informed decision without knowing.
Confidentiality: Lawyers must ensure the security of AI systems to prevent inadvertent or unauthorized disclosure of confidential client information.
Prevention of Misconduct, Including Discrimination: Lawyers must avoid misconduct, including dishonesty, fraud, deceit, misrepresentation, and discrimination, when using AI.
Oversight: Law firms and lawyers are responsible for ensuring the ethical use of AI by other lawyers and nonlawyer staff.
New York
The New York State Bar Association established a Task Force on AI and issued a 92 page Report and Recommendations in April 2024. Interestingly, the New York guidance notes an attorney’s duty to understand the benefits, not just the risks, of AI in providing competent and ethical legal representation and allows the use of AI tools to be considered in the reasonableness of attorney fees.
Competence and Diligence: A lawyer must understand the benefits, risks, and ethical implications associated with AI, including its use for communication, advertising, research, legal writing and investigation. Lawyers should consider whether AI tools will aid effective client representation.
Scope Of Representation: Lawyers should consider including a statement in engagement letters for acknowledgement by client that AI tools may be used.
Communication: A lawyer may use AI to aid communication but must still maintain direct and effective communication with a client not solely based on AI tools.
Fees: Effective use of AI tools can be considered in whether or not fees charged were reasonable.
Confidentiality: Lawyers must take precautions to protect sensitive client data and maintain client confidentiality, even with client’s consent to use confidential information in AI tools.
Supervision: A lawyer is still responsible for supervising the accuracy of the work produced by AI tools and the ethical use of AI tools by subordinate lawyers and non-lawyers.
Professional Independence and Unauthorized Practice of Law: AI is not a lawyer. It should always be subject to the professional judgment of a licensed lawyer.
Advertising and Solicitation: A lawyer must ensure that all content for advertising and solicitation purposes complies with ethical guidelines, including content generated using AI tools. Content should not be deceptive and AI may not be used to automatically generate phone calls, chat board posts, or similar solicitations.
As GenAI continues to develop, attorneys must navigate the intersection of technology and ethics carefully. The existing professional conduct rules provide a foundation for this, emphasizing competence, confidentiality, supervision, and the secure use of technology. While state-specific guidance on AI in legal practice is still evolving, the principles outlined in the ABA Model Rules and the proactive steps some states have taken regarding legal technology offer a roadmap for attorneys to follow.
Staying informed and engaged with both technological advancements and ethical standards will be key for lawyers aiming to responsibly harness the power of AI in their legal practice.
