If you’re wondering whether your law firm is getting calls driven by AI search results, it is.
The more interesting question is what’s already happened by the time the phone rings.. They described their situation to ChatGPT. They got a confident answer about their options, their timeline, what their case might be worth. Then they picked up the phone.
The legal industry gave CallRail one of the earliest signals of this. Across nearly 30 million inbound calls analyzed since November 2025, AI-referred calls to lawyers and law firms climbed through the first half of the year — part of a wider industry trend that's still in motion.
Legal's share of AI-driven calls climbed from late 2025 to spring 2026
AI referred 0.073% of inbound calls in November across every industry. Seven months later, that figure is now 0.115%, up 58%. The dip from May to June was small, and mostly reflects a significant increase in overall call volume, not a drop in demand.
The legal industry's own curve fills in some detail behind that opening number.
It led through March at 0.188% and April at 0.192%, then kept climbing to 0.194% in May before easing to 0.163% in June, its lowest point since tracking began.
While June’s figures interrupt legal's rise, a single month isn't enough to call it a lasting reversal. Each industry's percentage is measured on its own, so one industry's number going up doesn't cause another's to go down.
A more likely explanation is that seasonal demand is changing which industries rank highest from month to month.
Why legal services saw this happen before other industries
A person asking an AI assistant about a car accident or a custody question wants to feel confident in their situation before they call a law firm.
Asking for a recommendation like that is a different kind of question than asking an AI tool to compare products or explain a concept. It's part of why legal services consistently land near the top of this list.
Legal shares something else with the other industries near the top of this list, like manufacturing and home services. All three depend on expertise most people don't have on their own, and an AI assistant's value there is building confidence before that first call.
Prospects who researched their case move faster
AI-directed callers tend to skip much of the traditional browsing phase. There's no comparison shopping across several open tabs or filling out a contact form and waiting for a callback. The path runs straight from question to recommendation to phone call, often in one sitting.
Where speed-to-lead can decide who gets the case, that shorter path works in your favor, as long as someone picks up.
A caller who spent time researching their situation with an AI assistant, then picked up the phone, has already made up their mind. They're not calling to find out if they have a case. They're calling to find a lawyer.
What we are seeing from our CallRail attribution reports is the collapse of the sales funnel as we know it. It's forcing us to rethink the buyer's journey as AI-directed callers have higher sales intent and move faster when compared to traditional inbound leads.
– Will Troup, CEO of LocalSink
AI-informed clients arrive believing they understand their case
Not every AI-informed caller arrives believing they have the full picture. Some have done genuine research and come prepared with that information. Others have a single ChatGPT answer they never question — and they're just as confident.
Over the decades, I've welcomed — even invited — client input and discussion on important decisions. I think it's good for client relations, and it forces me to better refine my thinking. In the last three months, however, I've encountered something brand new: clients pushing back on strategy and advice based on ChatGPT. And the weird thing is, they seemingly believe ChatGPT more than my 25 years of experience.
– Attorney, Reddit
Others arrive holding onto a wrong filing deadline, an inflated case value, or an assumption about what a lawyer can do that doesn't hold up once the specifics come out.
Never in my career have I had my lawyering and my competency questioned as much by clients or their families as I have over the past few months, and all over things which are black and white, obviously legally incorrect. Nope, not going to file that motion based on hallucinated case law that appears to contain emojis.
– Public Defender, Reddit
Intake work now looks different, even when straightening out a misunderstanding only takes a minute.
Firms are now changing how they prep for these calls
Firms that keep their practice areas, service area, and hours easy for a large language model (LLM) to describe accurately give these tools less room to guess for prospects looking for legal representation.
On the intake side, some are training staff to ask early whether a prospect has already researched their situation with AI, so intake can start from the right questions instead of correcting course midway through. Some are also changing their paperwork.
I've started adding this into my fee agreements:
Please be advised that the complexity and scope of legal services may increase if you use third-party tools, including artificial intelligence (AI) platforms such as ChatGPT, to generate documents, advice, or analyses related to your matter. In such cases, the time required for us to review, verify, interpret, and supplement the AI-generated content may exceed initial expectations, and may result in additional charges beyond our original estimate.
– Attorney, Reddit
It's too early to know whether these calls are turning into signed cases or paying clients.
Whisper Messaging can flag that a call is AI-referred before staff even pick up, so intake can treat it differently from a standard call from the first greeting. Firms that get this right can skip the wasted questions and get straight to what the caller needs.
Know where your AI-referred calls are coming from
Without a way to trace each call back to its source, firms can't tell whether a new client found them through a billboard, a Google ad, or a conversation with ChatGPT — which means AI-driven calls are likely landing in reports as “direct” or “other,” uncredited and uncounted.
Call Tracking can help. More than 220,000 businesses already have access to CallRail's AI search attribution to see when a call traces back to an AI referral. Firms watching the data can spot which calls will need a few extra minutes to set the record straight.
See it for yourself, and put your marketing budget where it's working.
