Do lawyers have to tell clients when they use AI?
Ruling current. Adjudicated 20 October 2026; next review by 18 January 2027. 4 of 4 panel models answered.
Not legal or professional advice. This page is informational, not legal, medical or financial advice, nor any other professional advice. It is not a substitute for advice from a qualified professional about your own situation.
No blanket rule in the ABA Model Rules or most U.S. jurisdictions requires a lawyer to disclose AI use simply because AI is involved; disclosure becomes necessary only when the use is material to the client's matter, affects billing, implicates confidentiality, or is required by a court.
Agreed by all four
- There is no universal rule in the United States requiring lawyers to tell clients every time they use AI.
- Lawyers must be competent (ABA Model Rule 1.1), which includes understanding the benefits and risks of AI tools used in the representation.
- Lawyers must keep clients reasonably informed (ABA Model Rule 1.4); if AI use significantly affects the client's case, strategy, or outcome, disclosure may be required.
- If confidential client information is entered into a third-party AI system, the lawyer must protect that data under ABA Model Rule 1.6, and disclosure or client consent may be needed if the tool poses meaningful confidentiality risks.
- Billing must be honest and reasonable (ABA Model Rule 1.5); a lawyer cannot bill for hours of work that AI performed in minutes, and the fee structure should accurately reflect AI-assisted efficiency.
- A growing number of courts require attorneys to certify or disclose that AI-generated text in court filings has been reviewed for accuracy, this is a court-imposed requirement, not a general client-disclosure rule.
- The lawyer remains fully responsible for reviewing, verifying, and vouching for all AI-generated work product.
By three
- State bar associations, including those in California, Florida, and New York, have issued ethics opinions or advisories on AI that emphasize understanding its risks but stop short of requiring blanket client disclosure.
- Best practice, recommended by legal ethics experts, is to address significant AI use in the engagement agreement and obtain informed consent when confidential information or substantive legal work is involved.
By two
- Consumer-grade AI tools (such as the public version of ChatGPT) may retain user data for model training, which can violate client confidentiality and attorney-client privilege; lawyers must use secure, enterprise-grade legal AI software that guarantees data privacy.
- Clients have the right to ask whether AI is being used and how their confidential information is protected; an ethical lawyer should welcome and answer that question transparently.
Raised by one
- Several high-profile cases have already resulted in sanctions and public embarrassment after lawyers submitted AI-generated briefs containing fabricated case citations that were never verified.
- Texas has joined California, Florida, and other states in issuing formal ethics opinions on AI use in legal practice.
- Some courts and judges are moving toward flat-fee arrangements for AI-assisted tasks to avoid billing disputes over hourly rates.
WHAT WAS ASKED
Whether lawyers are legally or ethically required to disclose AI use to their clients. The question is unqualified, no jurisdiction specified, no context about what kind of AI use.
WHERE THE PANEL DISAGREED
There is no substantive disagreement on the governing rule. Every response reaches the same conclusion: no blanket disclosure requirement exists, but specific circumstances can trigger one. The differences are presentational. One response emphasizes client empowerment and the right to ask. Another leads with billing fraud risk. Another focuses on confidentiality as the sharpest edge. None of these represent a genuine disagreement, they are emphasis choices around the same core position.
One position worth examining: the framing that disclosure is simply "best practice" rather than ethically compelled in material cases. This softens what the ethics rules actually say. ABA Model Rule 1.4 does not merely recommend communication, it requires it when a reasonable client would want to know. If AI is performing substantive legal work, "best practice" understates the obligation. That softer framing does not survive scrutiny. The obligation is real where the use is material.
HOW STRONG THE AGREEMENT IS
The agreement is strong but sits on a moving foundation. The consensus holds as of mid-2026, no jurisdiction has yet enacted a blanket AI disclosure rule. What would overturn it is specific: a state bar issuing a formal ethics opinion requiring affirmative disclosure, or a court extending existing rules to mandate it by name. That is already happening in pieces, court filing certification requirements are spreading, and some bar ethics opinions are tightening. The consensus is correct today and directionally unstable.
The panel shares a common source set, ABA Model Rules, state bar opinions, and commentary from legal ethics scholars. This creates a shared blind spot: the question of what clients actually understand about AI use, and whether the "reasonable client" standard is being applied to a client population that largely does not know AI is involved in their legal representation. No response examined that gap seriously. The strongest counter-argument is that the existing "reasonable client" standard is underenforced, that if clients knew AI was drafting their motions or analyzing their cases, most would want to know, which means the existing rules already require more disclosure than lawyers are currently providing. That argument holds up. It is not resolved by the consensus.
WHAT THIS MEANS
The question has a clear answer for its general framing, and a harder answer when pressed.
The general answer: no blanket disclosure rule exists anywhere in the U.S. as of now.
The harder answer: the existing rules already require disclosure in more situations than most lawyers acknowledge. The trigger is not "AI was used", it is whether a reasonable client would want to know. For AI performing substantive legal work, handling confidential information in third-party systems, or affecting billing, the existing rules apply now. The gap is not a gap in the rules. It is a gap in compliance.
Courts are closing that gap from the other direction. Filing certification requirements are spreading. The trajectory is toward more required disclosure, not less. The current answer will age.
BOTTOM LINE
Lawyers are not required to disclose every instance of AI use. No jurisdiction has a blanket rule.
The decisive factor: existing communication and confidentiality rules already require disclosure when AI use is material, when it touches confidential data, affects billing, or substantially shapes the legal work. The absence of a specific AI disclosure rule does not mean the obligation is absent. It means it is already embedded in rules that existed before AI did.
What would change this verdict: a formal ethics opinion from any major state bar explicitly requiring disclosure across all substantive AI use, or a court extending filing certification requirements to client engagement letters.
The actionable directive: if you are a client, you have the right to ask, and a competent, ethical lawyer will answer directly. If you are a lawyer, "no rule requires it" is not a safe harbor when the use is material. The existing rules reach you.
Both blocks are SimulQ’s result exactly as a member receives it: four frontier models answered independently, and SimulQ’s protocol compared and ruled on their answers. Nothing above has been edited.
Sources the panel cited
- 2024 Formal Ethics Opinion 1 | North Carolina State Bar, ncbar.gov. Checked 20 October 2026.
- Ethics & Professional Responsibility - Artificial Intelligence - Website at University of Washington Law Library, lib.law.uw.edu. Checked 20 October 2026.
- Artificial Intelligence FAQs, michbar.org. Checked 20 October 2026.
Re-adjudication
Every Docket question goes back to the panel at least every 90 days. If the ruling changes, the change is recorded here.
- 20 October 2026. Re-adjudicated and republished as SimulQ's complete result (On the Surface and Digging Deeper), replacing the earlier summary.