Canada made a structural move in June 2026 that most edtech coverage under-read: the Association of Faculties of Medicine of Canada partnered with ScholarRx to advance open educational resources across all 19 Canadian faculties of medicine, built on modular English- and French-language "Bricks", shareable, editable units of curriculum content. The significance is architectural, a national bet on open, institutionally governed teaching infrastructure rather than nineteen separate procurement relationships with proprietary libraries, and it clarifies a question the whole market has been blurring: where institutional infrastructure ends and student-directed tools begin.
What open educational infrastructure changes
Three shifts follow from the model. Cost and equity: openly licensed core content lowers the floor for every faculty and every student, which matters in a bilingual system where French-language medical education has historically had thinner commercial coverage, and the EN/FR modularity is the partnership's most distinctive promise. Faculty agency: Bricks are editable, so a faculty can localise, provincial pathways, population-health emphasis, Indigenous health contexts, rather than teach around a fixed commercial product's assumptions, which is precisely the localisation work this series keeps finding at the centre of educational quality. And an AI authoring layer with governance attached: ScholarRx's faculty-facing tools, its TAI direction, sit inside institutional review rather than consumer generation, which is the difference between AI drafting content that faculty validate and students consuming generation raw, the validation-pipeline distinction at institutional scale: /blog/ai-draft-to-exam-ready-question-validation-pipeline.
Open resources versus proprietary libraries, honestly
The comparison is functional, not moral. Open infrastructure wins on cost, adaptability, language equity and institutional control, and carries the open-content burdens: maintenance depends on contributor energy, currency requires governance that must be resourced, and quality varies with the editing culture each faculty builds. Proprietary libraries win on production polish, integrated tooling and single-vendor accountability, and carry their own burdens: price, localisation limits, and the lock-in geometry treated at /blog/medical-ai-lock-in-who-owns-learning-history, now at institutional rather than individual scale. The mature reading is layered coexistence: open Bricks as the shared curricular spine, proprietary and independent tools where their components genuinely add, which is also the honest frame for where learner-directed platforms sit.
Where student-directed tools fit in the new landscape
The partnership covers institutional teaching content; it does not attempt the learner-side machinery this series decomposes elsewhere, blueprint-mapped question practice, adaptive targeting, spaced scheduling, Socratic remediation, examination calibration against MCCQE and beyond, and that division is healthy rather than competitive. A Canadian student's stack in the open-infrastructure era looks like: faculty Bricks as the taught spine; institution-governed AI where the faculty provides it; and independent learner tools, iatroX among the options for MCCQE-track calibration, general learning modes for dialogue, chosen by the year's binding problem per /blog/best-ai-stack-every-year-medical-school. For faculties, the strategic point runs the other way: open infrastructure raises the floor of content access, which moves the differentiating questions to exactly the layers procurement rubrics should already probe, educational design, update governance, jurisdiction fidelity, /blog/how-medical-schools-evaluate-ai-education-vendor. Canada has, in effect, nationalised the cheapest layer of the stack; the interesting competition now happens above it.
Frequently asked questions
Does this make commercial content obsolete for Canadian students?
No: it makes undifferentiated content harder to charge for, which is different; components open infrastructure does not supply, calibrated question banks, adaptive retention machinery, examination mocks, keep their value case intact.
How does the bilingual commitment change things practically?
Materially, if maintained: parity-quality French Bricks would reduce the translation burden francophone programmes have carried, with the clinical-nuance cautions of /blog/multilingual-medical-ai-translation-not-localisation applying to any AI-assisted authoring in both languages.
Should other countries copy the model?
The transferable part is the architecture, national coordination, open licensing, editable modularity, governed AI authoring; the enabling condition is a body like AFMC with convening power across every faculty, which not every system has.
What should students do with Bricks content in their own study?
Treat it as the taught spine it is: comprehension and localised context from the faculty's edited modules, with the retrieval, spacing and calibration layers still yours to run, since open infrastructure changes who owns the content, not how learning is manufactured.
How will quality be maintained across nineteen faculties?
The model's governance question, honestly open: editing rights need review cultures, currency needs owners, and the partnership's success will be measured in maintenance discipline rather than launch coverage, the same update-governance test this series applies to every content system.
Could the Bricks model extend to assessment content?
Technically yes and more carefully: shared formative items are a natural extension, while anything summative inherits the full validation-pipeline burden, blueprinting, psychometrics, security, which open collaboration can support but never skip.
Does open infrastructure change the equity picture for IMGs entering Canada?
Helpfully at the margins: openly accessible curricular content lowers one barrier to understanding what Canadian programmes actually teach, while the examination-calibration and licensure layers remain exactly as demanding, and exactly as separate.