ABA Formal Opinion 512 made it clear: you have a duty to communicate AI use to clients. But it didn't give you the letter. It didn't give you the engagement language. It didn't tell you exactly what to say when a client asks 'are you using AI on my case?'

That's what this is. A practical disclosure framework with sample language you can adapt today. Because the firms that handle AI disclosure well build trust. The firms that get caught using AI without disclosing it lose clients and face bar complaints — and the bar complaints are already happening.


What ABA Opinion 512 Actually Requires

ABA Formal Opinion 512 (July 2024) established that attorneys using generative AI must consider their duties under Rules 1.1 (competence), 1.4 (communication), 1.5 (fees), 1.6 (confidentiality), and 5.3 (supervision). On disclosure specifically, the opinion says attorneys must communicate with clients about AI use when it's material to the representation. But 'material' is doing a lot of work in that sentence. In practice, the safer interpretation is: if AI meaningfully contributed to work product, research, analysis, or strategy, disclose it. The risk of over-disclosing is zero. The risk of under-disclosing is a bar complaint. Several state bars — including California, Florida, and New York — have issued supplemental guidance that's more prescriptive. Check your jurisdiction's specific requirements.

The cleanest approach is building AI disclosure into your engagement letter. Here's adaptable language: '[Firm Name] may use artificial intelligence tools in connection with the services provided under this agreement. These tools may be used for legal research, document analysis, draft preparation, and other tasks where AI can improve efficiency and accuracy. All AI-generated work product is reviewed, verified, and supervised by licensed attorneys. Client confidential information is only processed through enterprise AI tools that meet our data security and confidentiality requirements and do not use client data for model training. The use of AI does not reduce the quality of legal services or the attorney's professional responsibility for work product. If you have questions about our AI use or wish to restrict it, please contact [attorney name].' This language covers communication, confidentiality, supervision, and provides an opt-out mechanism. Adapt it to your practice area and jurisdiction.

The Mid-Matter Disclosure: When AI Use Wasn't in the Engagement Letter

If you started using AI after the engagement letter was signed, or if your engagement letter doesn't address it, you need a supplemental disclosure. Keep it straightforward: 'I'm writing to inform you that our firm has begun using AI-assisted tools for [specific tasks: legal research, document review, contract analysis]. These tools help us work more efficiently on your behalf. Key points: All AI-assisted work is reviewed and verified by the attorneys assigned to your matter. Your confidential information is processed only through secure, enterprise-grade tools that do not use client data for training purposes. Our fees reflect the actual time spent on your matter, including any efficiency gains from AI use. Your rights and the quality of our representation are not affected. Please let me know if you have any questions or concerns about this approach.' Send this as a letter or email, keep a copy in the file, and note the client's response.

Addressing Client Concerns: The Conversations You'll Actually Have

Clients will ask three questions. 'Is my information safe?' Answer: We only use enterprise AI tools with strict data handling agreements. Your information is encrypted, not used for training, and subject to the same confidentiality protections as all client data. 'Am I paying for a computer to do the work?' Answer: You're paying for attorney expertise and judgment. AI helps us work more efficiently, which means faster turnaround and, in many cases, lower costs for you. Every AI-assisted work product is reviewed and verified by a licensed attorney. 'Can I opt out?' Answer: Absolutely. If you prefer we don't use AI tools on your matter, we'll honor that preference. Be aware that opting out may result in longer turnaround times and potentially higher costs for certain tasks. Be honest. Clients who feel informed and respected will almost always consent. Clients who discover AI use after the fact will almost always feel deceived.

Court-Specific Disclosure: The Rules You Can't Ignore

As of 2026, over 30 federal district courts and several state courts require specific disclosure of AI use in filed documents. The requirements vary — some require a certification that AI wasn't used for legal research, some require disclosure of specific tools used, and some require an attorney attestation that all citations have been verified. Your disclosure obligations to the court are separate from your disclosure obligations to clients. Build a pre-filing checklist that includes checking the specific court's AI disclosure rules. The Standing Orders page of each court's website is your source — and these rules change frequently. A disclosure letter to your client doesn't satisfy a court's disclosure requirement, and vice versa. Track both separately.

The Bottom Line: AI disclosure isn't a burden — it's a competitive advantage. Clients who trust you with transparency will stay with you longer. Courts that see you're taking AI seriously will give you more credibility. And the bar will never sanction you for being too transparent about your tools and processes. Build disclosure into your engagement letters, communicate proactively when practices change, and treat client questions as opportunities to demonstrate professionalism.

AI-Assisted Research. This piece was researched and written with AI assistance, reviewed and edited by Manu Ayala. For deeper takes and the perspective behind the research, follow me on LinkedIn or email me directly.