LegalVIS

How clients use AI to find a solicitor

People facing legal problems are no longer just searching Google. They are opening ChatGPT and describing their situation as if talking to someone they trust. That is already changing which law firms get found, and which do not.

Published by LegalVIS ยท Updated June 2026

When something goes wrong in a person's life, the first instinct is rarely to pick up the phone. It is to try to understand what is happening. A letter arrives from a landlord, or a spouse, or an employer, and the immediate response is not action but comprehension. What does this mean? How serious is it? What happens next? These are not questions people reach for a solicitor to answer. They are questions people reach for a friend.

For a long time, Google was the closest thing most people had to that friend in a legal crisis. Not because it was good at answering the question, but because it was available, private, and did not require explaining yourself to a stranger. You could type something half-formed and embarrassing at midnight and no one would know. The problem was that Google gave you pages, not answers. It gave you a list of places to go rather than anything that felt like guidance. You still had to do the work of reading, judging and deciding, at the moment when you felt least equipped to do any of those things.

What has happened in the past two years is that a meaningful number of people have discovered that ChatGPT does something different. It does not return a list. It listens. You explain your situation in the same language you would use with a family member, and it responds to the specifics of what you have said. It asks clarifying questions. It explains what the legal terminology actually means. It tells you what usually happens in situations like yours. And then, when you have been through all of that, it tells you who to call.

Legal problems sit in a category that researchers and search engines both recognise as uniquely high-stakes. Google formally defines this territory as Your Money or Your Life, or YMYL: topics where inaccurate information does not just mislead but causes genuine harm to a person's finances, safety, family or future. You can read how Google defines this in its own

. The legal sector sits squarely within it.

What that means in practice is that the person searching is not browsing out of curiosity. They are searching because something has happened and they need to understand it. The emotional state that accompanies that kind of search is a combination of anxiety, urgency and, often, a degree of shame. People feel embarrassed about not knowing things they feel they should know. They feel vulnerable about the situation they are in. They do not want to ask a real person, even a solicitor, until they feel they understand enough to ask the right questions.

This is the gap that AI has moved into, and it is not a small one.

Previsible's analysis of 1.96 million LLM sessions

What trust looks like at 11pm on a Tuesday

It is worth being specific about what this behaviour actually looks like, because it is easy to abstract it into a trend and lose sight of the human moment it represents.

A woman finds out her partner has been hiding debt and wants to know whether she is liable for it. She does not call a solicitor at that moment. She opens ChatGPT and types something close to what she would say if she were telling a friend. The AI responds to her specifically. It explains joint liability, asks whether they have joint accounts, outlines what she might want to document. By the end of the conversation, she has a clearer picture of her situation than she did before, and she has been given the name of a type of solicitor to look for and, in some cases, a specific firm.

A man receives a settlement agreement from an employer and has 10 days to respond. He has never dealt with anything like this before. He opens ChatGPT, pastes in the key clauses, and asks what they mean. The AI walks him through the standard terms, flags the ones that look unusual, tells him what a solicitor specialising in employment law would typically advise in this situation. And then it names firms.

In both cases, the person arriving at that recommendation has already been through a process that built trust. They did not receive a cold list of options. They received something that felt like help, and at the end of it, they received a name. The firms on that list are the ones that exist, in the AI's understanding of the world, as credible, verifiable and relevant. The firms that are not on it had no part in what just happened.

Previsible State of AI Discovery Report 2026

LegalVIS 2026 UK Law Firm AI Visibility Benchmark Report

, 335 firms, 11,879 pages crawled, 1,450 AI prompt tests.

The LegalVIS 2026 benchmark tested 335 UK law firms across 1,450 AI prompts. Of those firms, 51.7% were invisible across every single test run. They did not appear in a single AI response, on any platform, for any query connected to their practice areas. Not once (source:

LegalVIS 2026 UK Law Firm AI Visibility Benchmark

That figure is worth sitting with. More than half of the UK law firms tested do not exist, as far as AI systems are concerned. They have decades of experience, satisfied clients, expertise in their fields. None of that reaches the person on the other side of the conversation, because there is no record of it that AI systems can find, verify and use.

The 95.1% figure is equally telling. Of the firms that do appear in AI responses, almost all of them are actively recommended. The question AI search poses to a law firm is therefore binary in a way that traditional search never quite was. You are either in the conversation or you are not. There is no version of appearing on page two.

Why legal grew faster than any other high-stakes category

The reason legal is outpacing finance and healthcare in AI search adoption is not primarily about technology. It is about the nature of the need.

Financial questions, while serious, often have a degree of ongoing engagement. People check their pensions periodically, review their insurance annually, think about their investments over time. Healthcare questions are urgent but often public: people search for symptoms, for conditions, for medications, and they search repeatedly as situations develop. Legal problems tend to arrive suddenly, feel acutely personal, carry significant social weight, and require explanation before they require action. They are the category of problem most likely to make a person feel alone in what they are facing.

AI systems are particularly well suited to that kind of moment. They do not judge. They do not require you to already understand the vocabulary of the situation you are in. They respond to what you actually said rather than what you should have said. For a category defined by anxiety and vulnerability, that matters enormously.

"When a client in a high-stakes moment asks AI who to trust, you need to already be the answer."

When a client in a high-stakes moment asks AI who to trust, you need to already be the answer.

The firms that benefit from this are the ones that have made themselves legible to AI systems in advance of the moment the client arrives. That means structured profiles in legal directories that AI systems can verify, editorial mentions and citations in legal publications that confirm the firm's expertise in specific areas, correctly implemented schema markup on practice area pages, and a consistent identity across every source an AI might consult.

LegalVIS builds that infrastructure through

, which tells a firm exactly where it stands.

For a worked example of what that infrastructure delivers for a UK regional firm, read the

Find out whether your firm appears when a client needs you most.

Do clients actually use AI to find solicitors?

Yes, and the evidence points to the legal sector being particularly affected.

Why do people trust AI with something this serious?

Because the alternative, typing a half-formed legal question into a search engine at midnight and reading through a list of websites, does not actually answer the question. AI systems engage with the specifics of what a person has described, explain what it means, and then give a recommendation. That process builds trust through the conversation itself, before any firm is named.

Which AI platforms should law firms be concerned about?

ChatGPT and Gemini carry the highest volume of legal queries in the UK. Perplexity is growing and is particularly citation-driven, which means firms with published editorial mentions have a structural advantage on that platform. All three reward different types of credibility signals and need to be approached accordingly.

What does it mean for a firm to be invisible to AI?

It means the firm does not appear in AI responses to queries about its practice areas, regardless of the firm's actual quality, experience or reputation.

Where does a law firm start if it wants to improve its AI visibility?