How Law Firms Should Respond to Clients Using AI
Dealing with clients who present AI-generated legal work at their initial consultations is undoubtedly frustrating
Artificial intelligence (AI) has become the go-to source of information for virtually any topic, so it’s no surprise that people are using tools like ChatGPT for legal questions. Before meeting with a lawyer, people want to accumulate as much knowledge as possible about their legal issue and the challenges lying ahead, rather than feeling powerless. In fact, many people use AI to determine if they even need to hire a lawyer, or if they can just let technology guide them through their issue.
Dealing with clients who present AI-generated legal work at their initial consultations is undoubtedly frustrating. But as lawyers, it’s our responsibility to do what’s right for our clients, even those seeking legal advice from AI.
Communicate the Risks Directly and Without Condescension
Lawyers are well-aware of the dangers of using general purpose AI tools for specific legal questions, but most of your prospective clients don’t know what AI can’t do. And that being said, they may not be aware of what it can do – in the negative sense of the term.
General-purpose AI tools hallucinate. They cite cases that don’t exist, reference statutes that have been amended or repealed, and present all of it with equal confidence. OpenAI has explicitly updated its policies to state that ChatGPT should not be used for legal advice, but clients aren’t reading those policies.
AI also doesn’t understand context. It can’t weigh which details are legally relevant to a specific client’s situation, account for jurisdictional nuance, or recognize when an exception applies. The result is advice that sounds clear and actionable but is often misleading - sometimes dangerously so.
Explaining this once, clearly and respectfully, does two things: it protects the client, and it starts establishing why your judgment is worth paying for.
Set Clear Boundaries Around What You’ll Engage With When it Comes to AI
When clients present AI-generated research to lawyers, they often forget that you now have to spend more time explaining the flaws in their research. Time costs money, and every second you spend debunking AI-generated research takes time away from actual legal services.
At the same time, it’s unreasonable to think clients won’t consistently confer with AI before consulting with an attorney.
It’s entirely reasonable to establish in advance how you will respond to AI-generated documents as a starting point. How you’re willing to engage with specific questions the client has, based on what they read. That distinction matters. It keeps the conversation focused on the client’s actual situation, rather than a line-by-line critique of something a language model produced.
This isn’t dismissive. It’s a reasonable boundary that serves the client’s interests, and most clients will understand it when it’s explained in that way.
Use It as an Opportunity to Demonstrate Your Value
The silver lining in AI-assisted clients is that clients are arriving having thought about their problem. They have questions. They’re engaged. That’s a better starting point than a client who shows up with no preparation and no idea what they need.
The moment you take a question they brought from AI and give them an answer that’s specific to their situation, accounting for context, jurisdiction, relevant precedent, and the factors AI couldn’t weigh, you’re showing them the difference between a language model and a lawyer.
That is more persuasive than any explanation of why AI falls short.
Managing client expectations around AI is one piece of a larger shift in how law firms operate in this environment. If you’re thinking through how to position your firm, protect your time, and serve clients well, we’re here to help.
About the Author: Jim Field is the founder of Wellspring Business Strategies. An attorney and former CEO, Jim has spent over three decades leading complex operations across engineering and legal environments. He now works exclusively with law firms to improve operational efficiency, profitability, and long-term growth. His coaching philosophy is built on clarity, strategy, and execution.
Frequently Asked Questions: Responding to Clients Who Use AI
What should I do when a client brings AI-generated legal research to a consultation?
You don’t need to review or rebut the output line by line; that’s not the best use of your time or theirs. Instead, invite the client to identify specific questions they have based on what they read, and answer those questions in the context of their actual situation. That’s where your value is, and most clients will understand the distinction when it’s explained plainly.
Is it appropriate to tell clients not to use AI for legal research?
It’s appropriate to explain the risks clearly. General-purpose AI tools like ChatGPT are not designed for legal advice. OpenAI says so explicitly in its own policies. AI tools hallucinate, they miss jurisdictional nuance, and they can’t account for the specific facts of a client’s situation. You’re not telling clients what to do with their own time; you’re explaining why what they’ve brought to the table can’t substitute for professional legal judgment.
How do I explain the value of legal expertise to a client who thinks AI can do the same job?
Show them, don’t tell them. The most effective approach is to take a question they arrived with and give them an answer that accounts for their specific circumstances, the applicable jurisdiction, relevant exceptions, and strategic considerations. That experience is more persuasive than any abstract explanation of why AI falls short.
Does AI-generated legal research waste attorneys’ time?
It can, if you don’t set clear expectations in advance. The best approach is to establish early in the relationship what you will and won’t engage with. Client meetings should stay focused on their situation rather than on debunking AI output. That boundary is reasonable, it protects your time, and it serves the client better in the long run.
