Law firm AI search
Law firm AI search is the practice of prospective clients asking an AI system which lawyer to hire, and the answers those systems give. Joe Ashta works on it as a GEO / AEO consultant for law firms. This page describes what the answers look like, which kinds of firm are most exposed, what the published data shows, and what he does about it.
How prospective clients now ask
A person who needs a lawyer used to ask a friend, then search Google, then read a few websites. Many now go straight to ChatGPT, Claude or Gemini and ask in the words they would use with a friend: who is a good criminal defence lawyer near me, which firm should handle a wrongful dismissal, who does investor immigration. Google itself now answers first, through AI Overviews above the results and through AI Mode, its conversational tab. The same question reaches the same kind of system whichever door the buyer walks through.
What the answers look like
The answer is short. It names a few firms, usually with a sentence of reasoning for each: the practice area, a location, something the system found that suggests standing, such as years in practice, a notable matter, a peer rating or coverage in a publication. Some answers show source links; these are citations. The firm's name in the text is a mention. When the system tells the buyer to contact a firm, that is a recommendation.
Three things follow. The list is short, so most firms in any city are absent from it. The reasons are drawn from what the system can read about the firm, so the firm's public record decides the sentence written about it. And the buyer sees no ranking of the firms that were left out, so absence is silent.
Why referral-heavy practices are exposed
Firms that grow by referral have usually invested least in being findable. The site is a brochure, the lawyer profiles are thin, the directory listings were filled in once, and the firm's public record is whatever accumulated on its own. That was fine when the referrer was a person who already knew the firm.
An AI system is now a referrer, and it knows only what it can read. A firm with a strong reputation among lawyers and a weak public record is, to the system, a weak candidate. Referral-heavy practices are exposed because the thing that made them successful is invisible to the new referrer. Joe Ashta's BarTalk article, Beyond Referrals: How Criminal Defence Firms Are Found in the Age of AI Search, works through this for criminal defence, the practice area where the shift is easiest to see.
Big Law and consumer practices
The exposure takes two forms depending on the firm.
For Big Law, the questions are institutional. Which firms handle cross-border M&A in a given sector; which have a strong appellate practice; which lawyers hold particular credentials; which firm advised on a named transaction. The buyer is a general counsel or a procurement team, the query is about practice area and credentials, and the answer draws heavily on the firm's own publications, deal records and the legal press. The leaderboard segments (Big Law, Global, Mid-Law and Boutique) follow this side of the market.
For consumer practices, criminal defence, personal injury, family and immigration, the intent is local. The buyer wants a lawyer in their city, this week. The answer depends on whether the system can place the firm in that city with confidence, and on what third-party records say about it there. Local intent makes entity clarity the first problem, because a firm with an ambiguous name and address will lose to one the system can locate.
What the leaderboard shows
The BigLaw GEO Leaderboard, which Joe Ashta publishes, ranks 118 US law firms by AI search citations reported by Ahrefs across ChatGPT, Gemini, Google AI Mode and AI Overviews, refreshed monthly. Its general finding is that citation counts vary widely between firms of similar size and standing. Firms with comparable revenue, headcount and reputation are cited at very different rates. The leaderboard counts citations, which are a proxy. It does not measure which firms the platforms recommend, and it says nothing about any firm's legal ability. Domain Rating is shown beside each firm for context as a traditional SEO measure.
What Joe Ashta does about it
Joe Ashta measures which firms the AI platforms name and recommend for a firm's practice area and market, diagnoses why a firm is absent, fixes the causes (entity clarity, structured data, citation-worthy content, and corroboration from sources the firm does not control), and reports the result as a mention rate. The role is described in full on GEO consultant for law firms, and the person behind it on Joe Ashta. The canonical statement of both is the entity record. The previous page, AI search visibility, gives the mechanism the work above acts on.