Key takeaways
- Disclosure is required in some settings: Florida requires client-facing chatbots to say they are AI and not a lawyer[1], and Pennsylvania's joint opinion says lawyers must communicate with clients about AI use[2].
- Elsewhere it is a judgment call: New Jersey's rules do not require disclosure[3], while New York's City Bar says to consider disclosing AI use that is not routine[4].
- Clients rarely know: 71% of law firm clients do not know whether their firm uses generative AI[5].
- People want to be told: 76% of Americans say it is very or extremely important to know whether content was made by AI or a person[6].
Two audiences, two kinds of disclosure
Firms face the disclosure question in two places. The first is intake: a prospective client is talking to, or reading something produced by, an AI tool before any lawyer is involved. The second is the representation: a lawyer uses AI to research, draft or review work for a client. The rules and the risks differ, so decide each separately.
For intake, the question is identity: does the person know they are dealing with AI and not a lawyer? For representation, the question is informed decision-making: does the client know enough about how their matter is handled, what it costs and what happens to their information?
When disclosure is required
- Client-facing chatbots in Florida: Opinion 24-1 says chatbots that communicate with clients or third parties must include a disclaimer that the chatbot is an AI program and not a lawyer or employee of the firm[1].
- Clients of Pennsylvania lawyers: the joint bar opinion says lawyers must communicate with clients about their use of AI technologies, with clear explanations of how tools are used and their potential impact, and must disclose AI-related expenses[2].
- Confidential information and risky tools: ABA Formal Opinion 512 and California's guidance call for informed client consent before confidential information goes into tools that could expose it.
- Courts: some judges and courts require certifications or disclosure about AI use in filings.
- Client instructions: many corporate clients' outside counsel guidelines require notice or approval before AI is used on their matters.
When it is a judgment call
In many settings no rule requires disclosure. New Jersey's Supreme Court guidelines say the Rules of Professional Conduct do not require a lawyer to disclose the use of AI, while making clear that AI is no excuse for false or misleading content[3]. The New York City Bar takes a middle position: a lawyer should consider disclosing the intent to use generative AI that is not generally understood to be routinely used by lawyers, particularly when explaining fees and disbursements[4].
Routine is the useful word. Clients do not expect to be told that a lawyer used spell-check or a research database. They may expect to be told that an AI tool drafted the first version of their contract, reviewed their document production or summarized their deposition, especially if it changes the bill.
Why disclosing is usually wise anyway
Clients are mostly in the dark. 71% of law firm clients surveyed by Thomson Reuters did not know whether their firms used generative AI tools[5]. Members of the public care about the answer: 76% of Americans say it is extremely or very important to be able to tell whether something was made by AI or by people, and 53% are not confident they could tell on their own[6][7]. Half of US adults say they are more concerned than excited about the growing use of AI in daily life[8].
- Don't know whether their firm uses generative AI 71%
Source: Thomson Reuters Institute, 2025[5]
Show the data
| Item | Value |
|---|---|
| Don't know whether their firm uses generative AI | 71% |
Put together, silence is a risk. A client who later learns that AI was used, without having been told, is likely to wonder what else they were not told. A short, accurate disclosure made early costs little and protects trust.
A decision table
| Situation | Required? | Recommended action |
|---|---|---|
| AI answers calls or chats with prospects | Required in Florida for chatbots; wise everywhere | Disclose in the first sentence that the assistant is AI and not a lawyer |
| AI drafts or reviews a significant part of client work | Required in Pennsylvania; judgment elsewhere | Engagement letter paragraph plus a matter-level conversation |
| Confidential information goes into a tool with data risks | Informed consent expected | Explain the risk specifically and record consent |
| AI use changes fees or passes on a tool cost | Fee communication rules apply | Put it in the fee agreement before it happens |
| Routine features such as spell-check or search | Not required | No special disclosure |
| Court filing in a court with an AI order | Required by that court | Follow the order and keep the certification |
Corporate clients set their own rules
Business clients increasingly answer the disclosure question for you. Outside counsel guidelines from corporate legal departments and insurers often say whether AI may be used on their matters, which tools are acceptable, whether the client must approve first and how AI-assisted work may be billed. Those terms are part of the engagement, and breaching them is a client relations problem even where no ethics rule is involved.
- Read the AI clause of every set of outside counsel guidelines before staffing a matter.
- Record each client's position in the matter intake form so the whole team sees it.
- Where guidelines are silent, use your standard engagement letter paragraph and ask whether the client has a policy.
Prospects who never become clients
People who contact the firm and are declined still deserve accurate information about AI. If an AI tool handled their first call or chat, the non-engagement letter or closing message is a good place to repeat what happened to their information: that the conversation was recorded or transcribed, who at the firm can access it, and that it is kept confidential and deleted on the firm's schedule. That answer also helps if a prospect later asks why they spoke to an assistant rather than a person.
Wording you can adapt
Disclosures work best when they are short, specific and true. Avoid vague reassurance and avoid overclaiming.
- Phone or chat greeting: “Hi, you've reached the firm's AI assistant. I'm not a lawyer and can't give legal advice, but I can take your details and set up a call with our team. This conversation is recorded.”
- Engagement letter: “We use software tools, including AI tools, to help with research, drafting and document review. A lawyer reviews all work before it is relied on or sent. We do not put your confidential information into tools that would allow it to be used by others. We bill only for time our people actually spend.”
- Matter conversation: “For your document review, we plan to use an AI tool to sort the documents first, then our team will review what it flags. It should reduce cost. Here is how we protect your files. Do you have any concerns?”
- Website FAQ: one plain answer to “Does the firm use AI?” matching what the engagement letter says.
If an SI Employee such as Ava answers your calls, the greeting your firm approves can say exactly this: that Ava is the firm's AI assistant, not a lawyer, and that the call is recorded.
What not to say
- Claims that AI output is always accurate or checked by AI alone.
- Human names, photos or voices presented as staff for an AI assistant.
- Promises about speed or outcomes the firm could not put in an advertisement.
- Vague phrases such as “we use the latest technology” that tell the client nothing.
Advertising rules apply to anything the firm says publicly about its AI use, so keep website and marketing claims factual and consistent with what clients are told in engagement letters.
Example: rewriting a firm's disclosures
Example: a fictional estate-planning firm in Pittsburgh uses an AI drafting tool for first drafts of wills and trusts and an AI assistant on its after-hours phone line. Its engagement letter said nothing about AI, and the phone assistant introduced itself with a human first name. Following Pennsylvania's opinion, the firm adds the engagement letter paragraph above, has the phone assistant open by saying it is the firm's AI assistant and not a lawyer, and adds a website FAQ answer. Lawyers mention the drafting tool in the first meeting and note the conversation in the file.
Putting disclosure into practice
Map AI touchpoints
List every place AI touches prospects or client work: phones, chat, email drafting, research, review.
Check your states
Note which states' rules or opinions require disclosure for each touchpoint, and which courts you appear in have AI orders.
Write three texts
An intake greeting, an engagement letter paragraph and a website FAQ answer, all consistent.
Train lawyers
Lawyers know when a matter-level conversation is needed and how to record consent.
Review yearly
Update wording when tools or guidance change, and keep each version with its date.
For the underlying opinions, see ABA Formal Opinion 512, New York guidance and New Jersey and Pennsylvania.
Frequently asked questions
Do lawyers have to tell clients they use AI?
It depends on the state and the use. Pennsylvania's joint opinion says lawyers must communicate with clients about AI use; New Jersey's rules do not require disclosure; New York's City Bar says to consider disclosing non-routine use. Informed consent is expected before confidential information goes into risky tools.
Do AI chatbots on law firm websites need a disclaimer?
In Florida, yes: Opinion 24-1 requires a disclaimer that the chatbot is an AI program and not a lawyer or employee. Elsewhere it is strongly advisable.
Where should AI disclosure go?
In the first sentence of any AI conversation with the public, in the engagement letter for clients, in matter-level conversations where AI use is significant, and in a consistent website FAQ answer.
Should we disclose AI use if it lowers the bill?
Yes, it is a good reason to. Fee communication rules apply, and explaining how AI affects cost helps clients understand the bill.
References
The Florida Bar (2024). Florida Bar Ethics Opinion 24-1 (opens in a new tab) ↑
Pennsylvania Bar Association and Philadelphia Bar Association (copy hosted by LawNext) (2024). Joint Formal Opinion 2024-200: Ethical Issues Regarding the Use of Artificial Intelligence (opens in a new tab) ↑
New Jersey Courts (Supreme Court of New Jersey) (2024). Notice to the Bar: Legal Practice: Preliminary Guidelines on the Use of Artificial Intelligence by New Jersey Lawyers (opens in a new tab) ↑
New York City Bar Association, Committee on Professional Ethics (2024). Formal Opinion 2024-5: Ethical Obligations of Lawyers and Law Firms Relating to the Use of Generative Artificial Intelligence in the Practice of Law (opens in a new tab) ↑
Thomson Reuters Institute (2025). 2025 GenAI report: Executive summary for legal professionals (opens in a new tab) ↑
Pew Research Center (2025). How Americans View AI and Its Impact on Human Abilities, Society (opens in a new tab) ↑
Pew Research Center (2025). How Americans View AI and Its Impact on Human Abilities, Society (opens in a new tab) ↑
Pew Research Center (2025). How Americans View AI and Its Impact on Human Abilities, Society (opens in a new tab) ↑