Lawyers in Toronto that AI assistants recommend
We asked ChatGPT and Claude the four questions a buyer asks when looking for lawyers in Toronto, three times each, in English, and counted who they named. These are the businesses named in two or more of the 24 answers.
Names are printed as the assistants wrote them. We have not checked that any of them is good, open, or real.
- Answers
- 24
- Businesses named
- 42
- Named only once
- 20
- Named by 2+ assistants
- 15 of 22
- Named again on a rerun
- 50%
Edition September 2026. Download the data
Asked for lawyers in Toronto, ChatGPT and Claude most often named Epstein Cole LLP (25% of answers), McLeish Orlando LLP (25%) and Goodmans LLP (25%). Across 24 answers they named 42 different businesses, 20 of them only once. Each of four buyer questions was asked three times, in English, in September 2026.
The ranking
Share is the fraction of the 24 answers that named a business. The range under it is a 95% confidence interval: where two ranges overlap, the gap between those businesses could be chance.
| # | Business | Share | ChatGPT | Claude | Questions | Avg. position |
|---|---|---|---|---|---|---|
| 1 | Epstein Cole LLP | 95%: 12 to 45% | 25% | 25% | 1 of 4 | 1.5 |
| 2 | McLeish Orlando LLP | 95%: 12 to 45% | 25% | 25% | 1 of 4 | 1.7 |
| 3 | Goodmans LLP | 95%: 12 to 45% | 25% | 25% | 1 of 4 | 2.3 |
| 4 | Torys LLP | 95%: 12 to 45% | 25% | 25% | 1 of 4 | 2.3 |
| 5 | Miller Thomson LLP | 95%: 9 to 41% | 25% | 17% | 1 of 4 | 2.2 |
| 6 | Oatley Vigmond LLP | 95%: 9 to 41% | 25% | 17% | 1 of 4 | 2.2 |
| 7 | Blake, Cassels & Graydon LLP | 95%: 9 to 41% | 17% | 25% | 1 of 4 | 2.8 |
| 8 | Borden Ladner Gervais LLP | 95%: 9 to 41% | 25% | 17% | 2 of 4 | 3 |
| 9 | Nathens, Siegel LLPAlso written Nathens Siegel LLP | 95%: 7 to 36% | 17% | 17% | 1 of 4 | 2.3 |
| 10 | Fasken | 95%: 7 to 36% | 8% | 25% | 1 of 4 | 2.8 |
| 11 | Loopstra Nixon LLP | 95%: 4 to 31% | 8% | 17% | 1 of 4 | 1.7 |
| 12 | Thomson Rogers Lawyers | 95%: 4 to 31% | 8% | 17% | 1 of 4 | 3 |
| 13 | Norton Rose Fulbright Canada LLP | 95%: 4 to 31% | 25% | 0% | 2 of 4 | 3.7 |
| 14 | Davies Ward Phillips & Vineberg LLP | 95%: 4 to 31% | 17% | 8% | 1 of 4 | 4.3 |
| 15 | Dentons Canada LLP | 95%: 2 to 26% | 17% | 0% | 1 of 4 | 2 |
| 16 | Aird & Berlis LLP | 95%: 2 to 26% | 8% | 8% | 1 of 4 | 2.5 |
| 17 | Fitzgerald Law Group LLP | 95%: 2 to 26% | 17% | 0% | 1 of 4 | 3 |
| 18 | Klein Lawyers | 95%: 2 to 26% | 8% | 8% | 1 of 4 | 3 |
| 19 | Himel Farquharson Thorne LLP | 95%: 2 to 26% | 17% | 0% | 1 of 4 | 3.5 |
| 20 | Russell Alexander Collaborative Family Lawyers | 95%: 2 to 26% | 0% | 17% | 1 of 4 | 3.5 |
| 21 | Fasken Martineau DuMoulin LLP | 95%: 2 to 26% | 17% | 0% | 1 of 4 | 4.5 |
| 22 | McCarthy Tétrault LLP | 95%: 2 to 26% | 0% | 17% | 1 of 4 | 5 |
Each assistant column is the share of that assistant's own answers. Questions counts how many of the four produced the name at least once; position is where it came in the answer's list, 1 being first. 20 more names came up in a single answer and are counted above but not listed.
What each assistant said
The same questions, the same number of times, and often different names. Each column is one assistant's top five from the ranking above, with the share of its own answers that named them.
| # | ChatGPT | Claude |
|---|---|---|
| 1 | Epstein Cole LLP25% | Epstein Cole LLP25% |
| 2 | McLeish Orlando LLP25% | McLeish Orlando LLP25% |
| 3 | Goodmans LLP25% | Goodmans LLP25% |
| 4 | Torys LLP25% | Torys LLP25% |
| 5 | Miller Thomson LLP25% | Blake, Cassels & Graydon LLP25% |
Named here? Put it on your site
Every business in the ranking was named in two or more answers. The badge says that and nothing more: no position, because the ranges above do not support one. It links back to this page, so anyone can check.
<a href="https://aigrow.gg/index/lawyers/toronto"><img src="https://aigrow.gg/badges/index/2026-09/lawyers/toronto.svg" width="312" height="100" alt="Named by AI assistants: Lawyers in Toronto, AIGROW AI Visibility Index, September 2026"></a>The questions
Asked word for word, three times each, to every assistant through its API. The prompt is the one the free scan uses. No web search, no location and no chat history, so every answer starts from the same place.
- 1Who is the best lawyer in Toronto?
- 2I'm getting divorced in Toronto. Which family lawyer would you recommend?
- 3I was injured in a car accident in Toronto. Which personal injury lawyer should I call?
- 4Which law firm in Toronto would you recommend for a small business?
How to read this page
- What it measures
- Which names these models produce when asked cold, from what they learned in training. It is not a review, a rating or our recommendation, and a business missing here may simply be one the models were never shown.
- Why the ranges matter
- 24 answers is a sample. Asked again next month, most shares will move by a few points; the ranges say how far they could move by chance alone.
- One business, several spellings
- Branches, legal suffixes and spellings of one business are merged, and the other spellings are printed under its name. Names that might be two businesses are kept apart.
- A name is wrong
- If a business here is closed, misnamed or yours and you want it removed, tell us. We correct or remove it when the page is next published.
Elsewhere in the Index
Is your business on this list?
The free scan asks the same assistants the questions your customers ask, scores whether they name you, and lists who they name instead.
Edition September 2026 · data under CC BY 4.0