Making Sense.

Scotiabank launched Scotia Intelligence on April 13, a unified enterprise AI platform that allows them to deliver secure AI at scale for employees globally. 40% of their contact-centre queries and 90% of commercial email triage are already running through it. Scotia is the first of the Big Six to make a clean public statement about an integrated AI platform, but it won't be the only one. RBC, TD, BMO, CIBC, and National Bank are all in motion.

Anthropic released its most powerful model yet, Claude Mythos, gated to a small group of 50 organizations through Project Glasswing. The reason: AI can now find security flaws in software faster than most teams can fix them. OpenAI is also preparing customers for that future with a plan for AI-powered cyber defense.

For most Canadian organizations, the practical question is whether your defences can keep up with the new pace. Three things your IT team should pressure-test this quarter: how quickly your systems get updated when a vendor ships a fix, how fast you'd hear from a supplier if they were compromised, and the last time your team actually rehearsed an incident. The capability is gated today; it won't be a year from now.

The money is flowing towards efforts to position Canada within an AI-first world. Ottawa opened up their $890M Sovereign AI Compute Infrastructure Program, BC Hydro continues their competitive process for allocating clean power to AI projects, and Bell's AI Fabric will stand up their Groq-operated Kamloops facility in June. Cohere's $20B acquisition of Aleph Alpha also closed late last month, creating a transatlantic sovereign-AI player headquartered in Toronto and Berlin with endorsement from both Canadian and German governments.

If your procurement requires (or will soon require) Canadian data residency, your ability to host AI workloads in Canada will become more realistic in the near future.