Blog / 1 September 2026 · 5 min read
Should Your Firm Tell Clients When You're Using AI on Their Matter?
Most clients want to know when a law firm uses AI on their case, and most firms aren't telling them. Here's what the trust gap actually costs, and how to close it.
Quick Answer
Yes. 85% of clients say law firms should disclose when AI is used on their matters, yet most firms currently say nothing at all. This isn't a small preference gap, it's becoming a factor in client trust and retention. Firms that stay quiet about AI use, especially after a security incident, are seen as indifferent rather than careful. A short, proactive disclosure policy costs almost nothing to implement and directly addresses what clients are already asking for.
Why This Matters More Than It Seems
It's tempting to treat AI disclosure as a minor operational detail. The data suggests otherwise. Client trust in law firm technology is becoming a real factor in retention and revenue growth, not just a background consideration, and one in three clients say they want zero AI involved in their legal communications at all. Whether or not a firm agrees with that preference, it reflects genuine client sentiment that firms are largely not addressing.
At the same time, AI use itself is accelerating. Active generative AI use in legal organizations nearly doubled in a single year, and 60% of law firms are already deploying AI across practice areas. So the gap isn't shrinking. Firms are using more AI while saying less about it.
The Trust Gap in the Data
Clients want transparency, not necessarily less AI. The most consistent finding across recent industry research isn't that clients reject AI outright, it's that they want to know where it's being used, how it's managed, and where human judgment stays central. Clients are asking for clarity and confidence, not reinvention.
Most firms currently offer neither. More than half of clients say their firm has never proactively communicated about cybersecurity, and by extension, about AI use generally. This silence isn't neutral. In a relationship driven industry where client data is the core asset being managed, saying nothing tends to read as indifference rather than discretion.
The stakes are rising alongside adoption. 63% of firms experienced a significant email based security breach in the past year. Combined with rising AI use and low disclosure rates, this creates exactly the kind of situation where a client discovers AI involvement after something goes wrong, rather than being told upfront, which is far more damaging to trust than routine disclosure would have been.
What Proactive Disclosure Actually Looks Like
This doesn't require an elaborate client facing AI policy document, though larger firms may eventually want one. At minimum, effective disclosure usually includes:
- A brief, standard line in engagement letters noting that AI tools may be used to support research, drafting, or administrative tasks, with human review before anything reaches the client
- A short answer ready for any client who asks directly, rather than an uncomfortable pause
- Clarity that AI supports the work rather than replaces attorney judgment, since that's the specific reassurance clients seem to want most
- Consistency across the firm, so the answer doesn't depend on which lawyer a client happens to ask
What This Means for Smaller Firms
This trend has largely been discussed in the context of large firms and formal governance programs, but the underlying client expectation doesn't scale down. A client at a two person firm wants the same clarity as a client at a 200 person firm, arguably more, since smaller firms often have less formal process to point to when asked. A simple, honest disclosure practice is one of the lowest cost, highest trust building moves available to any size firm.
FAQ
Does disclosing AI use make clients trust a firm less?
The data suggests the opposite risk is larger, staying silent and having a client find out later, especially after an incident, does more damage than routine disclosure ever would. Clients broadly want clarity, not concealment.
What if a client says they don't want any AI used on their matter?
Given that a meaningful share of clients feel this way, it's worth having a real answer ready, whether that means adjusting workflow for that specific matter or explaining clearly where AI is and isn't involved. Avoiding the question isn't a sustainable long term answer.
Is this only relevant after a security incident?
No, waiting until after an incident is exactly the failure mode the data points to. Proactive, routine disclosure before anything goes wrong is what actually builds trust, not damage control after the fact.
The Bottom Line
The firms most exposed here aren't the ones using AI, most firms already are. They're the ones staying quiet about it while client expectations for transparency keep rising. A short, honest disclosure practice closes most of that gap for very little effort.
Not sure how your firm's current AI usage and client communication would hold up to this expectation? Our AI Tools Assessment looks at your workflows and helps you build a clear, honest answer before a client has to ask for one.