Canada just closed the AI adoption gap with the US.

Canadian AI adoption tripled in two years — 6.1% in 2024, 12.2% in 2025, 19.2% in 2026. The pace is the story: doubling every 12 months, which is the growth curve of infrastructure crossing an adoption threshold, not a fad curve. The US data point sits almost exactly on top of the Canadian 2026 number.

The planning-to-adopt signal is doing the leading-indicator work. In Q3 2024, the current + planning combined pipeline was 16.7%. A year later it was 26.7%. The actual 2026 use number (19.2%) came in below what the 2024 pipeline predicted — some intent doesn't convert — but the shape holds. Expect the Q3 2026 planning table (September release) to point toward the next leg.

Information & Cultural leads both countries (Canada 42.3%, US 39.7%) — the sectors where knowledge work is the product move first. Notably, Canada is ahead of the US in Finance (40.4% vs 33.9%) and Retail (19.2% vs 14%). The US only publishes three sector cuts in the current BTOS wave; the Canadian data covers all 16 NAICS sectors.

US enterprise (250+ employees) leads at 37%, well above Canada's 100+ employee bucket at 27.8%. But the more interesting cut is Canada's 1–4 employee firms at 19.9% — matching the national average. Solopreneurs and micro-businesses in Canada are adopting AI at essentially the same rate as everyone else, which isn't how digital tech usually rolls out.

Data analytics (36.6%) and text analytics (34.5%) top the list — text-heavy, review-friendly workloads where humans still gate the output. LLMs specifically show up at 24.8%. Marketing automation using AI is a new 2026 category and already at 19.8% — that shift alone is worth watching quarter over quarter.

Worker-reported GenAI use (22%) is running slightly ahead of business-reported AI use (19.2%). Not a huge gap — but the direction is the interesting part. It says workers are adopting AI at least as fast as their employers are formally deploying it, likely faster in the shadow-AI band. Managers thinking their team isn't using AI yet are probably wrong.

The behavioral data reframes the adoption story. Canada doesn’t just adopt AI at roughly the US rate — Canadians use Claude at more than 4× their population share, the second-heaviest of any major country. And what drives it isn’t provincial income, it’s the share of professional, scientific, and technical services jobs. The translation signal is the most Canadian thing in the dataset: bilingualism policy showing up as a usage pattern, concentrated in the provinces that run the most public administration.

When someone quotes an AI adoption number — 19%, 50%, 78% — the question isn't “which one is right?” It's “which construct?” Ramp counts firms *paying* for AI. Atlanta Fed weights firms by employment (large firms drag the number way up). CSWC asks workers directly. CSBC asks businesses. Each of these is correct — for what it measures. The tracker leads with CSBC and BTOS because they're the most comparable representative frames.