AI Regulation Tracker / Legal profession and ethics
Florida Bar Opinion 24-1: Lawyers Can Use Generative AI, With Four Conditions
Issued January 19, 2024, it was one of the first state-bar ethics opinions on the technology and it still reads as a clean checklist. The Florida Bar did not ban generative AI for lawyers. It said you may use it, so long as you protect client confidentiality, verify the output, bill honestly, and keep your advertising compliant.
This one is older than most items on the tracker, and I am flagging that up front. Florida Bar Ethics Opinion 24-1 came out on January 19, 2024. I am covering it now because it is foundational. It was one of the first state-bar ethics opinions in the country to take generative AI head on, and the framework it laid down has aged well and keeps getting cited as later states line up behind the same logic. If you want to understand how the American legal profession decided to handle lawyer AI use, this is close to the source.
The headline is permissive. In the opinion's own framing, a lawyer "may use generative artificial intelligence ('AI') in the practice of law but must protect the confidentiality of client information, provide accurate and competent services, avoid improper billing practices, and comply with applicable restrictions on lawyer advertising." No ban, no special permission slip. Just your existing duties, applied to a new tool.
What does the opinion require on confidentiality?
This is the condition with the most teeth. The Bar tells lawyers to actually look under the hood of the tool before trusting it with client information, researching the program's policies on data retention, data sharing, and self-learning. The point is that some AI tools train on or expose what you type in, and putting confidential client information into a tool like that can be a disclosure. The opinion goes further and advises that a lawyer should obtain "the affected client's informed consent prior to utilizing a third-party generative AI tool if the utilization would involve the disclosure of confidential information." In plain terms: know how the tool handles data, and if using it means client confidences leave your control, get the client to say yes first.
How far does the verification duty go?
All the way. The Bar's move here is to treat generative AI like a nonlawyer assistant. You can delegate work to it, but you own the result and you have to supervise it. That means reviewing AI-generated work product the same way you would review a paralegal's, verifying the accuracy and sufficiency of anything it produces, including research. The reason is not theoretical. Generative AI fabricates citations and authority that look completely real, and a lawyer who files that work without checking it is the one on the hook. The opinion closes the escape hatch that using a third-party AI tool somehow shifts responsibility off the lawyer. It does not.
What are the billing rules?
The Bar is direct about not letting AI efficiency turn into a billing windfall or a hidden charge. You cannot double-bill, and you cannot bill a client for time you did not actually spend just because AI did the work faster. An AI subscription is overhead, not a cost you can prorate and pass through per client. And if you intend to charge a client the actual cost of using generative AI on their matter, you inform the client first, preferably in writing. The through-line is honesty: the client should not be surprised by an AI-related charge, and should not pay twice for the same work.
Why this matters beyond Florida
Two reasons. First, Opinion 24-1 is advisory. It is guidance interpreting the Florida Rules of Professional Conduct, not a rule or a discipline order by itself. But the duties it applies, confidentiality, competence, supervision, reasonable fees, and truthful advertising, are binding rules, so a lawyer who ignores the guidance is still exposed under those rules. Second, the structure travels. Florida was early, and a string of other states has since reached materially the same place: use AI if you want, but confidentiality, verification, and honest billing are non-negotiable. If you practice anywhere in the US and you want a short, sane statement of what your AI duties look like, this opinion is a good one to keep on the desk.
Questions professionals are asking
Does Florida Opinion 24-1 ban lawyers from using AI?
No. It permits generative AI in legal practice. It conditions that use on four existing duties: protecting client confidentiality, providing competent and verified work, billing honestly, and complying with advertising rules. The opinion is permissive, not prohibitive.
Do I need client consent to use a generative AI tool?
Where using a third-party tool would involve disclosing confidential client information, the opinion advises obtaining the affected client's informed consent first. It also tells lawyers to research the tool's data-retention, data-sharing, and self-learning policies before trusting it with client information.
Can I bill a client for using AI?
Not by double-billing or charging for time you did not spend. An AI subscription is overhead and should not be prorated as a per-client cost. If you plan to charge a client the actual cost of using AI on their matter, tell them first, preferably in writing.
Is the opinion binding, and does it matter outside Florida?
It is advisory guidance interpreting the Florida Rules of Professional Conduct, not a binding rule on its own, but those underlying rules are binding. Outside Florida it is influential: it was among the earliest state-bar AI opinions and other states have adopted materially similar confidentiality, verification, and billing expectations.
RELATED BRIEFINGS
Browse the full AI Regulation News tracker
Informational analysis for working professionals, not legal advice. Confirm how any ethics opinion or rule applies to your practice with your bar's ethics counsel in the relevant jurisdiction.