Florida Bar Ethics Opinion 24-1 as an Operations Checklist

On January 19, 2024, The Florida Bar issued Ethics Opinion 24-1 on lawyers' use of generative artificial intelligence. The opinion concludes that lawyers may use generative AI, but must protect client confidentiality, provide accurate and competent services, avoid improper billing practices, and comply with lawyer advertising restrictions. Like other Florida advisory ethics opinions, it states on its face that it is not binding.

LexAIAdvisors is not a law firm and does not give legal or ethics advice. What we can do is translate the opinion's four themes into the operational controls a small firm would need to show that it took them seriously. Treat what follows as a starting checklist, and confirm your own obligations with ethics counsel or The Florida Bar's ethics resources.

Confidentiality: know where the data goes

The opinion says lawyers must protect confidentiality by researching an AI program's policies on data retention, data sharing, and self-learning. It recommends obtaining a client's informed consent before using a third-party generative AI program if that use would involve disclosing confidential information, and notes that consent under the confidentiality rule is not required if the use does not involve disclosing confidential information to a third party. It also applies the Bar's earlier cloud-computing guidance to AI vendors: confirm the provider has an enforceable obligation to keep information confidential and secure, investigate the provider's reputation and security, and find out whether it retains or claims rights to what the lawyer submits.

Operationally, that becomes a vendor file for every AI tool the firm uses. It should hold the current terms of service and privacy policy, the answer to whether inputs are used to train models, the retention period for prompts and outputs, and the settings the firm has chosen, such as training opt-out and history retention. Personal consumer accounts used for client work are the most common gap. Firm-managed business accounts, with settings controlled centrally, close it. The engagement letter is a natural place to address client consent where the firm's lawyers decide it is needed.

Oversight: treat AI like an assistant whose work gets checked

The opinion notes that Rule 4-5.3 defines a nonlawyer assistant as a person, but treats that rule's standards as useful guidance: the firm should have policies that reasonably assure AI use is compatible with the lawyer's obligations, and lawyers must review AI work product as they would a paralegal's, including verifying the accuracy and sufficiency of any research. It also says a lawyer may not delegate to AI any act that could constitute the practice of law.

In practice that means a written AI use policy, even a one-page one, that lists approved tools, prohibited uses, and a required human review step before anything AI-assisted leaves the firm. Citation checking deserves its own step; our earlier post on Florida's citation-accuracy certification for court filings covers why. Keep a simple log of which matters used which tool, so the firm can answer questions later.

Fees and costs: efficiency cannot inflate time

The opinion states that efficiency gains from AI must not result in falsely inflated claims of time, and suggests that flat or contingent fees may let the benefit of efficiency accrue to lawyer and client alike. It says a lawyer should inform the client, preferably in writing, of an intent to charge the actual cost of using generative AI; that if the actual cost for a particular matter cannot be determined, the lawyer should not prorate the subscription and should instead treat it as overhead; and that time spent developing minimal competence with an AI tool should not be charged to the client.

The operations translation: decide in advance how AI subscriptions are booked (overhead unless a per-matter cost is genuinely traceable), make sure time entries describe the work actually performed, and keep training time off client bills. If the firm intends to pass through AI costs, the fee agreement template should say so.

Advertising and chatbots: say it is a bot

On advertising, the opinion says generative AI chatbots that communicate with clients or third parties must comply with advertising rules and must disclose that the user is communicating with an AI program, not a lawyer or law firm employee. It cautions against an overly welcoming chatbot that might give legal advice, fail to identify itself, or omit clear disclaimers limiting the lawyer's obligations, since that could create a prospective client or even a lawyer-client relationship without the lawyer's knowledge. It suggests screening questions to limit communication with people already represented by another lawyer, and says firms may advertise their use of AI but may not claim their AI is superior to others' unless the claim is objectively verifiable.

For a website chatbot, the checklist is: an AI disclosure in the first message, a clear disclaimer before any facts are collected, factual questions only with no advice, a represented-party screening question, and transcripts saved and reviewed by a person. Marketing copy should describe what the firm does with AI, not rank its tools against anyone else's.

The short version

Approved tools on business accounts, a vendor file for each tool, a one-page use policy with a mandatory human review step, billing rules decided before the first invoice, and a chatbot that identifies itself and does not advise. None of that requires new software. It requires someone to write it down and check it quarterly.

LexAIAdvisors helps small Florida firms build this layer through its AI agent workflow advisory: tool selection, data settings, review checkpoints, and intake bots designed around disclosure. If you want a gap assessment against the controls above, request an audit at /audit.