Florida AI Ethics Rules for Lawyers

The official instruments

An early opinion that aged well

Florida moved fast: Opinion 24-1 issued January 19, 2024, months before ABA Formal Opinion 512, and its four-part structure (confidentiality, oversight, fees, advertising) anticipated the national consensus closely enough that Florida has felt no need to revisit it. Two of its moves remain distinctive and worth understanding precisely.

First, the consent posture. Where some states locate consent at the open-model line (Oregon) or at exposure risk generally (the ABA), Florida frames informed client consent as the recommended course before confidential information is disclosed to a third-party generative AI system. In practice that makes Florida a consent-forward jurisdiction: a Florida firm whose workflow routinely places client information into external AI tools should treat consent language in the engagement letter as standard equipment rather than an exotic add-on. The mechanics, and how Florida compares to the written-consent requirement in West Virginia or the no-per-se-duty states, are mapped in the comparison table.

Second, the supervision model. Opinion 24-1 analogizes generative AI to a nonlawyer assistant: the lawyer may delegate work to it but must supervise it and remains responsible for its product, the same Rule 4-5.3 architecture that governs paralegals. North Carolina later adopted the same framing. The analogy does real work: it imports a complete, familiar compliance vocabulary (review before release, no independent judgment, responsibility stays with the lawyer) instead of inventing a new one, and it is the reason the verification workflow is not optional in Florida: you would not file a paralegal’s research unchecked.

The advertising section nobody else wrote

Because Florida’s advertising rules are the most elaborate in the country, 24-1 is the rare AI opinion with a developed advertising analysis, and it lands on the use case small firms actually deploy first: intake chatbots. The rules: a chatbot that screens or communicates with prospective clients must disclose it is an AI program and not a lawyer or firm employee; chatbot interactions must respect solicitation limits; and AI-generated marketing claims live under the same truthfulness rules as everything else Florida regulates heavily. For the many Florida firms buying AI intake from vendors, this is also a supervision point: the vendor’s chatbot is your nonlawyer assistant, and its script is your responsibility.

Billing, and the double-billing flag

Florida’s fee analysis matches the national line (actual time only; learning time is the lawyer’s own cost) with one emphasis other early opinions skipped: double-billing. AI’s speed makes it newly easy to bill two matters for one stretch of work, and 24-1 names the practice as the Rule 4-1.5 violation it is. Florida firms adopting AI should pair the efficiency conversation with timekeeping hygiene; the billing guide covers the fee-design alternatives that capture gains legitimately.

The Florida to-do list

Verify any AI tool’s terms before client information enters it, and put consent language in engagement letters if external tools are part of the workflow. Run every output through human review and independent citation verification before filing; Florida’s federal districts include judges with AI standing orders, so the per-filing forum check in the court orders guide applies. Configure intake chatbots to disclose their nature in the first interaction. Bill actual time, watch for double-billing, and keep tool-learning off client bills. And write it all down: the supervision analogy makes firm measures the unit of compliance, which is what the policy template provides out of the box.