Alex Tapscott's Canada Day op-ed turns a real, disruptive incident - Anthropic being forced to cut off foreign users of its own top models - into a case for treating AI infrastructure as sovereign national infrastructure, not a rented utility. For practitioners and policymakers outside the US, the underlying incident is the more durable takeaway: it is now a documented example of a foreign government's export-control decision reaching directly into which AI models a company's own non-US staff can use.
What happened
On June 12, 2026, a US export-control directive required Anthropic to suspend access to its most advanced models, Fable 5 and Mythos 5, for any foreign national anywhere, including Anthropic's own non-US employees. Because the company could not reliably distinguish users' nationality in real time, it disabled both models globally to comply. The US Commerce Department lifted the restriction on June 30, and Fable 5 resumed for global users while Mythos 5 access was restored for some US organizations following June 26 government approval.
Editorial analysis
Writing in BetaKit, Alex Tapscott, CEO of CMCC Global Capital Markets and co-founder of the Blockchain Research Institute (BRI), uses the incident as his central evidence for treating AI as national infrastructure rather than a rentable service. He draws an extended analogy to the 1885 completion of the Canadian Pacific Railway, arguing that just as Canadian-owned rail secured the country's economic sovereignty, Canada now needs its own compute, data, and identity infrastructure rather than depending on foreign-owned platforms it does not control. He proposes coordinating Crown land, hydroelectric capacity, and capital from Canada's large pension funds (the "Maple 8") and major banks to build domestic AI infrastructure, rather than having government build it directly.
Background
The op-ed is a companion piece to "Rebuilding Canada for the New Technology Order: A National Moonshot for a Prosperous and Sovereign Canada," a BRI whitepaper Tapscott co-authored with Don Tapscott and released June 16, 2026 with support from more than 50 business, government, and academic contributors. That report explicitly cites the Anthropic export-control directive as the immediate impetus for its sovereign-infrastructure recommendations, which span digital infrastructure, scaling domestic tech firms, productivity, defense, and a broader social contract for an AI economy.
What to watch
Whether Canadian federal or provincial governments act on sovereign-compute proposals (land, power allocation, procurement) rather than issuing statements alone; whether other countries cite the Anthropic episode to justify their own sovereign-AI or data-residency policies; and follow-on coverage of the BRI report's specific recommendations as they move toward policy proposals.
Key Points
- 1A US export-control directive briefly cut off Anthropic's Fable 5 and Mythos 5 for all non-US users, showing foreign AI access can vanish overnight.
- 2Tapscott cites that episode to argue Canada needs sovereign compute, data, and identity infrastructure, not just policy statements.
- 3The op-ed builds on a June 16 BRI report Tapscott co-authored, continuing momentum behind Canada's sovereign-AI policy debate.
Scoring Rationale
A well-documented real incident (a US export-control directive disabling Anthropic's own top models for its non-US staff) used as the evidentiary anchor for a serious sovereign-AI-infrastructure policy argument from a credentialed author tied to a substantive BRI whitepaper. Notable for practitioners and policymakers tracking AI export-control fallout and sovereign-compute debates, though it remains an opinion piece rather than an enacted policy.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems
