UK Law Firm AI Visibility Benchmark 2026

A sector-wide audit of AI visibility readiness across 335 UK law firms, published 1 June 2026 by LegalVIS. Each firm was scored against the LegalVIS 5 Pillar Audit System and tested against ChatGPT and Gemini for AI recommendation rates.

Headline findings

Pillar averages

A sector that performs well on structure and poorly on discovery. On Structural Readability the sector average is 66.7; on Entity and Topic Strength, 58.5. The problem is Pillar 1, AI Discoverability, where the sector averages well below 25.

Score distribution

The sector sits just below half marks, with a 46 out of 100 average and almost no firms approaching the top of the range. A third of all firms cluster in a single 10-point band, 41 to 50 out of 100. 71% of firms score 20 or below on Pillar 1; the vast majority have not begun to address AI discoverability.

Regional findings

Geography makes almost no difference. All four UK regions sit within 4 points of each other on overall AI readiness. Rest of UK leads Pillar 1 with 25.1, but every region scores critically low — confirming this is a sector-wide problem, not a regional one.

AI prompt test results

Audited UK law firms appear in 22.8% of ChatGPT responses and 23.3% of Gemini responses to relevant legal search prompts. Firms with strong trust signals (3 to 4 signal types) appear in AI responses at 59.8%, compared to 40.8% for firms with weak signals (0 to 1 types). 79.1% of UK law firms have no detectable case study signals.

Kingsley Napley appeared in 112 prompt tests, more than any other firm in the dataset. The top 20 firms appearing in AI responses were not part of the audited cohort; they are the firms AI systems default to when legal queries are run.

Methodology

335 UK law firms were audited via the LegalVIS 5 Pillar Audit System, an automated crawl across 11,879 pages covering AI Discoverability, Brand and Category Clarity, Entity and Topic Strength, Structural Readability, and Trust and Evidence Signals. AI recommendation rates were measured by running a standard set of legal search prompts against ChatGPT and Gemini and recording firm appearances.