
What Canadian privacy law actually requires of your AI
There is no federal AI statute in force in Canada. What binds you is privacy law — PIPEDA, BC PIPA, and Quebec's Law 25, which already regulates automated decisions.
Insights
Long-form notes on AI strategy, training, and implementation. These pieces are written like field notes from the work, not abstract thought leadership.

There is no federal AI statute in force in Canada. What binds you is privacy law — PIPEDA, BC PIPA, and Quebec's Law 25, which already regulates automated decisions.

AI-native is usually defined so that only startups qualify. Here is the version that works for a 30-year-old business — built on your people, not your org chart.

Only 13% of employees have had any AI training while 80% already use AI tools at work. That gap is where shadow AI lives — and why most training changes nothing.

Most readiness assessments produce a scorecard nobody acts on. A useful one answers five questions and tells you what not to do yet.

Most enterprise AI pilots never ship. The blocker is almost never the model — it is ownership, integration, and the day-two work nobody scopes.

Governance that reviews everything reviews nothing well. Risk-tiering, a named owner, and clear defaults let teams ship — and beat the August 2026 EU deadline.

Agents can hit 60% success on a single run and 25% across eight. Session dashboards hide that completely — here is what to measure instead.