Ethics in Brief —Behind the Brief: Do Clients Need to Know About AI Use in Legal Drafting?

By Alexis C. Garcia, Esq.

Many attorneys now rely on AI tools to draft pleadings, summarize materials, or evaluate legal theories before consulting Lexis and Westlaw. While its speed and convenience are undeniable, AI is no white knight—its outputs can be inaccurate or biased, client information can be compromised, and overreliance may erode a lawyer’s independent professional judgment. These risks place every day AI use squarely at the intersection of emerging technology and a lawyer’s ethical duties to their client. 

As AI becomes embedded in practice, the critical question shifts from whether lawyers may use these tools to how they must manage their use—and, in particular, whether clients must be informed when AI contributes to the work. Responsible AI use demands more than efficiency; it requires thoughtful oversight and adherence to ethical obligations. In practical terms, this means juggling the “5 C’s”: Competence, Confidentiality, Communications, Candor, and Charges—making sure every tool supports the client’s interests rather than cutting corners.

Competence

Rule 1.1 mandates that lawyers provide competent representation to clients. While they are not required to be experts in generative AI, lawyers must exercise the “skill, and mental, emotional, and physical ability reasonably necessary” for its use. The American Bar Association’s Formal Opinion 512—its first ethics guidance on a lawyer’s use of AI tools, released in July 2024—stated that, “[g]iven the fast-paced evolution of generative AI, technological competence requires lawyers to remain vigilant about both the benefits and risks of these tools.” In practice, this means carefully reviewing and refining AI-generated work, supplementing it with independent research, and applying critical analysis to ensure that every argument and reference fully reflects the lawyer’s professional judgment. (Rule 1.3.) 

Confidentiality

Although the use of AI is not inherently improper, a lawyer may not use client information in a manner that could “disadvantage” the client without first obtaining the client’s informed consent. (Rule 1.8.2.) Therefore, lawyers must avoid entering identifiable client information into AI tools that lack adequate confidentiality protections, must anonymize sensitive data where appropriate, and must carefully review the provider’s terms of use. (Rule 1.6.) Consulting cybersecurity professionals and implementing strict data retention protocols are no longer optional—they are essential to safeguarding client confidentiality.

California’s recent Senate Bill 574 (“SB 574”) underscores this obligation. The bill would prohibit attorneys from inputting confidential or personally identifiable client information into public AI systems, require verification of AI-generated outputs, and mandate measures to prevent bias or inaccuracy. By codifying these responsibilities, SB 574 highlights that careful handling of client data is not only a best practice but an emerging legal requirement, reinforcing the confidentiality duties under Rule 1.6.

Communications

Rule 1.4 requires that a lawyer reasonably consult with the client about the means by which the client’s objectives will be accomplished. Whether this duty requires disclosure of AI use depends on the facts of each case.Ultimately, the touchstone is informed decision-making. If the use of a particular AI tool implicates confidentiality risks, billing considerations, or strategic judgment in a way a reasonable client would consider significant, transparency is the prudent course. (Rule 1.2.) 

Candor 

Before submitting any AI-assisted work to a client or court, a lawyer must carefully review and verify all content—including analysis and citations—for accuracy, correcting any errors or misleading statements. (Rule 3.1.) In an era where courts are increasingly slapping sanctions on attorneys for citing “hallucinated” cases, lawyers should perform their due diligence and determine whether their jurisdiction expects full disclosure of AI use under local rules, standing orders, or other court directives. (Rule 3.3; see also Noland v. Land of the Free, L.P., B331918 (Sept. 12, 2025) [the court imposed $10,000 in monetary sanctions on plaintiff’s counsel for filing briefs containing fabricated legal citations generated by AI, directed counsel to serve the opinion on his client, and ordered the clerk to notify the State Bar].) 

To help lawyers navigate this evolving landscape, Ropes & Gray maintains a nationwide tracker of courts and judges with orders related to the use of AI, providing a practical resource for identifying jurisdictions with specific disclosure or verification requirements. (See e.g., Outreach v. Medoff, 2025 Cal. App. Unpub. LEXIS 8133 (Cal. Ct. App. Dec. 17, 2025) [the court imposed $5,070 in monetary sanctions payable to opposing counsel, ordered the opinion forwarded to the State Bar under Bus. & Prof. Code § 6086.7(a)(3), and required counsel to provide the opinion to her client and certify compliance]; Howell Mgmt. Servs., LLC v. Vashisht-Rota, No. D086055, 2025 Cal. App. Unpub. LEXIS 8006 (Cal. App. 4th Dec. 15, 2025) [the court imposed $59,235 in attorney fees payable to opposing counsel and $15,000 to the court clerk for violating California Rules of Court].) 

Charges

Faster drafting shouldn’t compromise fee transparency. According to the Practical Guidance For The Use Of Generative Artificial Intelligence In The Practice Of Law, lawyers may use AI tools to work more efficiently and can bill for the time actually spent crafting prompts, refining inputs, or reviewing outputs, but cannot charge for hours saved by AI. Any costs tied to AI must follow applicable law, and the fee agreement should clearly outline all fees and expenses, including those related to AI use. (Rule 1.5.) 

Accordingly, the integration of AI into legal practice offers a streamlined approach and an analytical boost, but it also amplifies ethical responsibilities. Lawyers cannot outsource judgment, confidentiality, or candor to a machine.By combining vigilance, transparency, and careful review, lawyers can harness AI responsibly, delivering the benefits of technology while upholding the standards their clients—and the law—demand.