Technology report, October 2026

About the data. The national and sector figures are Statistics Canada's own: the Canadian Survey on Business Conditions for AI use, second quarters of 2024 to 2026, and the Job Vacancy and Wage Survey for hiring (Statistics Canada, table 33-10-1167; table 14-10-0442). Our firm postings cover 61 law, accounting, consulting and engineering firms, collected October 10, 2026, and only six met our coverage bar, so they count as supporting evidence, not a finding. Intervals are 95 percent percentile bootstrap intervals from 2,000 resamples with a fixed seed.

Dataset. Job postings for 61 of the 63 Canadian law, accounting, consulting and engineering firms in our list, the largest by headcount plus the regional firms the audience outside Toronto and Montreal knows first, listed in our firm list with their source rankings (Harper Grey and Pitblado joined the list after collection and are covered from November). Postings come from each firm's careers page, from monthly Internet Archive snapshots of that page over the 12 months from October 2025 to October 2026, from Workday job feeds and from Job Bank Canada, the Government of Canada job board, de-duplicated by firm, title and link. Collection on October 10, 2026 produced 1,458 unique postings for 50 of the 61 firms (512 from Workday job feeds, 467 from 171 monthly archive snapshots, 436 from live careers pages, 43 from Job Bank). Only 35 firms could be checked in the Internet Archive before the archive stopped accepting connections. A firm counts as covered when at least 6 of the 12 months have archived postings and it has at least 20 unique postings: 6 firms meet that bar, with 793 postings between them, 28 of them AI roles. That is too few to stand as a headline, so the firm data is reported as supporting evidence only.

Source

Firm careers pages, Internet Archive snapshots, Job Bank Canada and Workday job feeds, collected October 10, 2026

Method

De-duplicated by firm, title and link; titles classified with the strong keyword list below

Flag: law headcount substitute. Law firm headcounts use the lawyer counts in Lexpert's Largest Firms list (https://www.lexpert.ca/canadas-largest-law-firms, retrieved October 10, 2026; the edition year is not stated on the page) in place of the Canadian Lawyer ranking named in our firm list. The Lexpert list covers 25 of the 29 law firms in the collection; the other four keep the headcount on file, unverified.

Source

Lexpert, "Canada's Largest Law Firms", retrieved October 10, 2026

Method

Lawyer counts as listed, matched to firms by name; edition year to confirm before publication

Sample. Twenty-nine law firms, fourteen accounting firms, ten consulting firms and eight engineering firms, 61 in all, each tagged national or regional; the list now holds 31 law firms and 63 in all. Firms whose careers sites returned no postings are reported as missing, not as zero: Borden Ladner Gervais, Blake Cassels & Graydon, McCarthy Tetrault, Stikeman Elliott, Dentons Canada, Cassels Brock & Blackwell, Aird & Berlis, Baker Tilly Canada, Stantec, Hatch and Mallette. Of the six covered firms, three (Osler, the accounting firm MNP, formerly Meyers Norris Penny, and Raymond Chabot Grant Thornton) posted no AI roles.

Classification. A posting is artificial intelligence (AI) related when its title matches one of: artificial intelligence, machine learning, generative AI, AI, large language model, data scientist, data science, natural language, prompt engineer, ML (machine learning) engineer, deep learning. Titles matching only "automation", "innovation" or "legal technology" are counted separately and reported in the note.

Correlation. AI share per firm against headcount (Spearman rank correlation with a 2,000-resample percentile bootstrap interval), practice mix, billing model and tier (Kruskal-Wallis test and share of variance between groups). Not computed for this issue: six firms meet the coverage bar, all with a headcount on file, and a rank correlation on six points would mean nothing; the test waits for at least ten covered firms.

Bimodality. Bimodality coefficient (above 0.555 suggests two modes) and a two-component Gaussian mixture against one component by the Bayesian information criterion (a gap above 10 in favour of two is strong evidence). Not run: the test needs at least ten covered firms and six qualify.

Benchmark. Statistics Canada, Canadian Survey on Business Conditions (CSBC), the quarterly federal survey of business conditions, second quarter of 2026 (table 33-10-1167, released May 27, 2026). In that wave, 19.2 percent of Canadian businesses said they had used AI to produce goods or deliver services in the previous 12 months, up from 12.2 percent in the second quarter of 2025 and 6.1 percent in the second quarter of 2024. In professional, scientific and technical services, the industry group that holds law, accounting, consulting and engineering, the rate was 32.4 percent, against 31.7 percent a year earlier and 13.7 percent in 2024 (Statistics Canada, table 33-10-1167; table 33-10-1004; table 33-10-0825). Among businesses with 100 or more employees it was 27.8 percent. In the third quarter of 2026 wave (released August 31, 2026), 25.2 percent of all businesses said they planned to use AI in the next 12 months, against 54.8 percent in professional, scientific and technical services (table 33-10-1207-01; quality grade B). By province, AI use ran from 13.1 percent in Manitoba and 13.6 percent in Saskatchewan to 25.1 percent in British Columbia; urban businesses were at 21.0 percent and rural ones at 9.9 percent; information and cultural industries were at 42.3 percent, finance and insurance at 40.4 percent and health care at 30.0 percent. The survey does not publish law or accounting separately, so the benchmark is the wider industry group. The survey reports shares of businesses, not a distribution of intensity, so it cannot by itself show a bimodal pattern; that test rests on our firm data.

Source

Statistics Canada, tables 33-10-1167 and 33-10-1004, full-table downloads through the Web Data Service (WDS), retrieved October 10, 2026; 2024 rates from table 33-10-0825-01 and planned use from table 33-10-1207-01, full-table downloads retrieved October 11, 2026. Quality grades: A (excellent) for the all-industry rates, B (very good) for the professional services and 100-plus employee rates.

Productivity check. Labour productivity by industry, from Statistics Canada table 36-10-0480 (real output per hour worked, business sector industries), set against each industry's AI use rate. The table is inactive and ends in 2024, so the check covers growth from 2023 to 2024 in the 13 industries that match between the two surveys. The rank correlation on 2024 use rates is -0.31 (95 percent interval -0.78 to 0.31); on 2026 use rates it is -0.20 (interval -0.73 to 0.44). It is reported in the argument as a supporting line, not as a finding.

Source

Statistics Canada, table 36-10-0480, retrieved October 12, 2026; tables 33-10-0825 and 33-10-1167

Method

Spearman's rank correlation with a 2,000-resample percentile bootstrap interval, fixed seed

Hiring demand. The Job Vacancy and Wage Survey (JVWS), the quarterly federal survey of unfilled positions, does not publish law, accounting, engineering or consulting separately; its finest cut is the three-digit professional services sub-sector. The industry group posted 36,730 job vacancies in the second quarter of 2026, against 37,020 a year earlier, and its vacancy rate was 2.9 percent against 3.1 percent across all industries; payroll employment was flat at about 1.23 million, while the average hourly wage offered for vacant positions rose 3.7 percent, from C$38.90 to C$40.35. Its occupation table has no industry split, so data scientist vacancies can only be read for the whole economy: 1,435 in the second quarter of 2026, against 1,250 a year earlier, out of 548,400 vacancies in all occupations. For scale, the Business Register counted 158,435 professional services businesses with employees in June 2026, of which 199 had 500 or more employees and 17,423 were offices of lawyers.

Source

Statistics Canada, tables 14-10-0442-01 (JVWS by industry, quarterly, unadjusted), 14-10-0444-01 (JVWS by occupation unit group, quarterly, unadjusted) and 33-10-1174-01 (Canadian Business Counts, with employees, June 2026), retrieved October 10, 2026

Method

Vacancies, vacancy rate, payroll employees and offered wage for North American Industry Classification System (NAICS) sector 54, second quarter of 2026 against second quarter of 2025; vacancies for National Occupational Classification (NOC) unit group 21211, Canada, all industries; business counts by employment size for NAICS sector 54 and its industries

What would change this finding. The next comparable survey wave, expected in the second quarter of 2027: a sector rate above 35 percent would mean the plateau was a pause, not a stall. Firm postings with at least ten covered firms that showed AI roles concentrated in a few large firms would turn a sector stall into a split between leaders and the rest.

Licence. Careers page facts and Job Bank listings are public; Statistics Canada Open Licence for the benchmark; firm filings are public records.

Sources by section. Each sentence in the issue links to its own source; these lines give the full citation and how each figure was produced.

Source

Statistics Canada, Canadian Survey on Business Conditions, tables 33-10-1167-01 (second quarter of 2026) and 33-10-1004-01 (second quarter of 2025); Job Vacancy and Wage Survey, table 14-10-0442-01 (quarterly, unadjusted); full-table downloads through the Web Data Service, retrieved October 10, 2026; 2024 rates from table 33-10-0825-01 (second quarter of 2024) and planned use from table 33-10-1207-01 (third quarter of 2026), retrieved October 11, 2026

Method

Share of businesses reporting AI use (and, for plans, planning AI use in the next 12 months) in producing goods or delivering services over the last 12 months (quality grade B for the industry group, A for all industries), second quarter against second quarter; vacancies, vacancy rate, payroll employees and offered wage for NAICS sector 54, second quarter of 2026 against second quarter of 2025

Source

Statistics Canada, tables 33-10-1167, 33-10-1004, 14-10-0442 and 14-10-0444, retrieved October 10, 2026; tables 33-10-0825-01 and 33-10-1207-01, retrieved October 11, 2026; full tables 33-10-0825, 33-10-1004 and 33-10-1167 by industry, province and urban or rural, and table 36-10-0480 (labour productivity by industry), retrieved October 12, 2026

Method

Ratios and changes computed from the published rates; the industry test is Spearman's rank correlation between each industry's AI use rate and its labour productivity growth, with a 2,000-resample percentile bootstrap interval (fixed seed); firm AI share is AI-titled postings over all unique postings for the six firms meeting the coverage bar

Source

Statistics Canada, tables 33-10-1167, 33-10-1004 and 14-10-0442, retrieved October 10, 2026

Method

Recommendation drawn from the flat adoption rate and the flat vacancy rate over the same period

Source

Statistics Canada, tables 33-10-1167 and 14-10-0442, retrieved October 10, 2026; our firm postings, collected October 10, 2026

Method

The 35 percent threshold is prediction tech-2026-10-2 in the predictions ledger; the ten-firm bar is the coverage rule in the Dataset note above

Source

Vendor price pages, retrieved October 10, 2026, OpenAI and Google rows re-verified October 11, 2026: https://platform.openai.com/docs/pricing (gpt-6.1-sol, standard tier, short context), https://docs.anthropic.com/en/docs/about-claude/pricing (base input and output tokens; Claude Sonnet 5.5 2.00 and 10.00, Claude Opus 5.5 4.00 and 20.00, verified October 10, 2026), https://ai.google.dev/gemini-api/docs/pricing (gemini-3.1-pro-preview, paid tier, standard, prompts up to 200K tokens; output includes thinking tokens)

Method

Prices in United States dollars as published; the cost per task will be the field test's token count multiplied by list price, from the November issue