Every number on this page is from a named, linked source, and every number was re-verified on August 23, 2026. The short version: 42% of U.S. consumers now say they would use ChatGPT to research a lawyer — up from 9% three years ago — Google's share of that research has fallen 15 points in a single year, and the pages AI engines cite are increasingly not the pages that rank. Here is the data, with the honest caveats attached.
How many people use AI to find a lawyer in 2026?
More than at any point measured, and the growth is accelerating. In iLawyerMarketing's August 2026 survey of 1,110 U.S. consumers, 41.9% said they would use ChatGPT to research which lawyer to hire — up from 28.1% in 2025, 21% in 2024, and 9% in 2023. Half of all consumers (50.1%) would use at least one AI answer engine, and roughly 1 in 10 (9.5%) would research a law firm entirely through AI — no Google, no directories. A year earlier, that AI-only consumer barely existed.
The pattern extends beyond legal. BrightLocal's 2026 Local Consumer Review Survey found 45% of consumers used AI to find a local business or service in the past year — up from 6% a year earlier — making AI the #3 local discovery channel, behind only Google and Facebook.
Two demographic findings matter for law firms specifically. The heaviest AI adopters are not the young: in the iLawyerMarketing data, consumers aged 45–60 lead every age group — 76% would use AI to research which firm to hire — and households earning $125,000+ adopt at the highest rates. The clients many firms most want are the ones most likely to be asking an AI first.
Is Google still where clients look for lawyers?
Yes — and saying otherwise would be wrong. Google remains the most-used source for researching attorneys at 71.9%, well ahead of ChatGPT's 41.9%. About 70% of ChatGPT users still also use Google. AI search is an additive shift, not a replacement.
But the direction is unambiguous. Google's share fell from 86.7% to 71.9% in one year — a nearly 15-point drop, the largest iLawyerMarketing has measured — and the share of ChatGPT users who double-check on Google fell from 94% to about 70% as consumer trust in AI answers rose (53.5% say they trust AI more than a year ago). The honest summary: Google is still the biggest surface, and a fast-growing share of client research now happens where only the named firms exist.
What happens to clicks when an AI answer appears?
Most of the click value evaporates for everyone except the sources the answer names. In Pew Research Center's analysis of 68,879 real Google searches, users who saw an AI summary clicked a traditional result on just 8% of visits, versus 15% without a summary — and clicked a source cited inside the summary only 1% of the time. Sessions were also more likely to simply end after an AI summary (26% vs 16%).
In legal specifically, a 2026 RepuClinic analysis reports law-firm website traffic fell roughly 19% in 2025 even where Google rankings held steady — rankings and results have decoupled. And legal informational queries — "how does a deposition work," "what is adjustment of status" — trigger Google AI Overviews at about 77.7%, the highest rate of any YMYL category (MeanPug Digital analysis, via Attorney at Law Magazine): the research questions that precede hiring a lawyer are exactly the ones AI now answers first. (Both are single-vendor analyses; we attribute them as such.)
Does ranking on Google put a firm in AI answers?
Less and less — this is the most important structural fact on this page. Ahrefs' study of 863,000 keywords and 4 million AI Overview citations found only ~38% of cited pages also rank in Google's top 10 for the same query — down from 76% in mid-2025. The remaining citations split roughly evenly between positions 11–100 and beyond position 100. Ranking well still helps: Hoverboard AI's analysis of 102,000 AI responses found a #1 Google ranking earns a brand mention 32–49% of the time depending on the engine. Both things are true at once: a top ranking improves your odds, and most of what AI engines cite never ranked at all. SEO alone no longer covers the surface where clients are asking.
How many firms does an AI answer actually name?
Very few — that is the winner-take-all dynamic. Citation-pattern studies (Profound; Qwairy's analysis of 118,000 answers) put ChatGPT at roughly 2.6–4.5 citations per answer — one or a few names, not a page of ten. Sources concentrate hard: analyses of ChatGPT citations find a small set of domains captures the large majority of all citations. And BrightLocal found ChatGPT currently recommends only about 1.2% of local business locations — nearly every firm is invisible on the new surface. The name an engine learns for a niche tends to keep being named; a lead here is a first-mover advantage you then defend, not a permanent moat.
Do small firms have a real chance in AI search?
Yes — and the budget data explains why the field is open. Per 2026 legal-marketing benchmarks, only ~14% of solo attorneys and ~32% of small firms even have a formal marketing budget; high-growth firms spend ~16.5% of revenue on marketing versus ~5% for no-growth firms. Most small firms are not competing for AI visibility at all. Meanwhile the big firm's decade of SEO compounding transfers only partially (the ~38% figure above), and AI engines reward being the clear specialist answer for a specific community, practice area, and place — not the biggest budget. On broad head terms, deep-rooted incumbents still win; in specific niches, no one's roots run deep yet.
What is being named in an AI answer worth?
Where it can be measured, citation carries measurable value: Seer Interactive's CTR research found pages cited in an AI Overview earn ~35% more organic clicks than equivalent uncited pages, and advertisers cited in the Overview see ~91% more paid clicks. For context on the paid alternative: personal-injury clicks run $150–$500+ in competitive metros, with documented extremes above $1,000 per click (2026 legal-CPC benchmarks). Visibility in the answer is the asset the market is repricing.
How should a law firm measure AI-search visibility?
Skeptically, and per engine. AI answers vary by engine, by time, and by phrasing, so a single screenshot proves little. Our standard — the one we apply to our own clients and publish — is a six-surface panel (ChatGPT · Perplexity · Google AI Overviews · Google AI Mode · Gemini · Claude), every question asked at least three times, depersonalized, with raw evidence saved so a skeptic can re-run it, reported per engine with no composite score, including where a firm is not visible. The full method is documented in the VERDICT framework and how we measure AI visibility.
Sources and verification
Every statistic above was re-verified against the linked source on August 23, 2026. These numbers move quickly; where a figure comes from a single vendor analysis rather than corroborated research, we say so inline. Sources: iLawyerMarketing 2026 consumer study (n=1,110, Aug 2026); Pew Research Center (Jul 2025); Ahrefs; BrightLocal LCRS 2026; Seer Interactive; Profound; RepuClinic; MeanPug Digital; MyCase/Grow Law 2026 legal-marketing benchmarks.
Market statistics describe the market, not any individual firm's results. AI-search visibility can be measured; rankings, leads, and case outcomes are never guaranteed (Cal. RPC 7.1). If you want to know where your own firm stands, that is a measurable question — the AI Visibility Audit answers it with evidence you keep.
