Spellbook Outranks Thomson Reuters in ChatGPT Legal Tech Ads
Spellbook leads ChatGPT legal tech at 23.1% share (July 2026), but the top 3 hold just 48.7%, and an 'Unknown' brand ties General Counsel AI for 4th.
As of July 2026, Spellbook leads legal tech on ChatGPT with 18 of 78 placements (23.1% share), outranking a fragmented field of 26 advertisers where the top 3 brands hold 48.7% combined. The leader is an AI-native contract review tool, not a legacy incumbent, and no single brand has locked the category yet.
An AI-native contract review tool now outranks the legacy giant Thomson Reuters on ChatGPT. Spellbook leads at 23.1% share while the incumbent sits tied for second, evidence that AI-native entrants are reshaping this category faster than legacy vendors can react.
Twenty of the 26 advertisers hold four or fewer placements. That is what an unowned surface looks like at the buyer level: a sprawl of single-digit presences with no brand claiming the prompt sets buyers are actually typing. The top of the leaderboard has one clear name, and the rest is open territory.
Spellbook's lead is structurally telling, not just large. Legal tech is the kind of category where a handful of legacy CLM vendors normally set the terms. On ChatGPT, that hierarchy is being rewritten by an AI-native entrant that did not exist in the prior distribution stack. The leader holds nearly double the volume of the next brand, and the next brand is tied with Thomson Reuters, the kind of incumbent that should be consolidating this surface. It is not.
- Spellbook23%
- Others22%
- Ironclad13%
- Thomson Reuters13%
- General Counsel AI, Inc.6%
- Others23%
An 'Unknown' entity is tied with General Counsel AI, Inc. for 4th largest presence in legal tech on ChatGPT, with 5 placements and 6.41% share each. That is not noise. It is a competitive blindspot, and with 26 advertisers already active, the share that is still claimable is shrinking monthly. Build a weekly prompt-monitoring habit now, or inherit the category narrative someone else is already writing.
Where to actually spend
The data points to a specific move: own a sub-cluster instead of the niche. The 'solo, attorney, practice management' cluster is the only sub-cluster captured in the data, with 7 advertisers splitting 12 placements and no clear winner. A challenger who concentrates spend on comparison prompts like 'best case management software for solo family law attorney 2026' can plausibly win the cluster, given that no brand has claimed it yet.
The same logic applies to campaign motion. Six named campaigns are running, each tied to a different buying intent: Ironclad on Enterprise, Conga on AMER-OAI-CLM-Clicks, General Counsel AI, Inc. on demo, BigID on AI governance and DSPM, and HoneyBook Inc on a generic tag. Match the campaign shape to the prompt's buying stage, not the other way around.
Keep reading
We probe ChatGPT with realistic consumer prompts and capture the sponsored ad cards it returns — every creative, the triggering prompt, and the advertiser. Figures reflect our captured sample, not OpenAI's internal data. Explore the live ad library and market intelligence.