AI Shortlist / LegalTech

in-house legal AI agents

Who ChatGPT, Perplexity, and Gemini name when B2B buyers ask about in-house legal ai agents.

12 prompts · 36 answers · Data as of Oct 27, 2026, 3:27 AM ET · Methodology

IS MY COMPANY ON THE SHORTLIST?

SHARE OF VOICE

Who AI recommends

  1. 1CoCounsel
    26/36 · 72%
  2. 24/36 · 67%
  3. 23/36 · 64%
  4. 14/36 · 39%
  5. 14/36 · 39%
  6. 13/36 · 36%
  7. 10/36 · 28%
  8. 8Thomson Reuters CoCounsel
    9/36 · 25%
  9. 6/36 · 17%
  10. 6/36 · 17%
  11. 11Lexis+ AI
    5/36 · 14%
  12. 12Lexis+ with Protégé
    5/36 · 14%
  13. 13Thomson Reuters CoCounsel Legal
    5/36 · 14%
  14. 14Confidentiality
    4/36 · 11%
  15. 15LexisNexis Protégé
    4/36 · 11%

Mentions out of 36 successful answers across all three engines.

BUYER PROMPTS

What we asked, ranked by buyer intent

These are the questions buyers type when they are close to choosing a tool, not when they are just learning. We rank them by how close the wording is to a purchase: price and budget questions first, then switching and "alternatives to" questions, then head-to-head comparisons, then "best tool for my kind of team". Each prompt was asked once in ChatGPT, Perplexity and Gemini, with no login, chat history or personalization. Full methodology

WHERE AI GETS ITS ANSWERS

Top cited domains

  1. 1lumay.ai104
  2. 2spellbook.com87
  3. 3legalontech.com69
  4. 4gc.ai40
  5. 5harvey.ai37
  6. 6legal.thomsonreuters.com29
  7. 7legora.com27
  8. 8thelegalprompts.com21
  9. 9lexisnexis.com20
  10. 10thomsonreuters.com19
  11. 11eve.legal18
  12. 12aivortex.io18

COMPARE

Head-to-head in this niche

Pick two brands to see mentions per engine and shared prompts.

GET THE FULL MAP

Get the full map plus a personalized plan for your company

Prefer to talk now? Book a 15 min call.

Ready to move? Book a 15 min call.

METHODOLOGY

  • Engines: ChatGPT, Perplexity, Gemini.
  • This niche: 12 prompts, 36 successful answers. Tested 2026-10-27.
  • A mention is one successful answer that names the brand. Share of voice is mentions divided by total successful answers in the niche.
  • Results are a point-in-time snapshot of model answers, not a traffic ranking.