Article · November 17, 2026 · Karolyn Law, Founder
Canada’s Privacy Regulators Just Wrote the Rules for Health AI. Clinics Should Read Them Now.
On October 29, 2026, Canada’s federal, provincial and territorial privacy commissioners and ombuds published a joint resolution with an unusual request: make AI governance in health care a legal requirement, not a selling point.
AnalysisThe resolution, Responsible use of artificial intelligence in health and health research, was adopted at the regulators’ annual meeting in Ottawa on September 15, 2026 and published on October 29. It calls on governments across Canada to review, amend or adopt legislation so that AI systems in health care meet six core principles, and to legally require everyone who develops, procures or uses those systems to establish effective AI governance and accountability frameworks.
The six principles are short enough to fit on an index card:
- Privacy protective
- Valid and reliable
- Safe
- Human rights affirming
- Transparent
- Accountable
As Privacy Commissioner of Canada Philippe Dufresne put it: “Privacy matters to Canadians – especially when it comes to sensitive health information.”
First, the honest status
A resolution is not a law. Nothing in it binds a clinic or a vendor today, and this article would not deserve your trust if it pretended otherwise. The obligations in force today come from existing privacy laws: PIPEDA federally and, in New Brunswick, the Personal Health Information Privacy and Access Act.
But regulators do not publish joint resolutions for decoration. The federal, provincial and territorial regulators adopted the same text, and they committed to coordinate, to the extent possible, on joint enforcement, guidance and public education behind it. When the people who run privacy oversight across the country agree on where the rules are going, procurement follows, because no clinic wants to buy a system it will have to replace when the law catches up. The resolution also sits in a pattern: Ontario’s Information and Privacy Commissioner published guidance and a checklist on AI scribes for the health sector in January 2026, and Ontario’s Auditor General reported in May 2026 on what happens when AI tools are approved without the assessments to back them up. We wrote about that report here: Ontario’s Auditor General tested 20 approved AI scribes.
What the resolution asks of clinics, in plain terms
The part of the resolution aimed at health care providers reads less like philosophy and more like a purchasing checklist. Paraphrased, it asks providers who adopt AI to:
- Designate a contact person patients can raise concerns or objections with about how AI uses their information, and give them a practical way to challenge consequential decisions or inferences the system made.
- Explain clearly how information will be collected, used and disclosed and, where consent is required and an opt-out is practicable, offer one before the system is deployed.
- Sign vendor contracts that limit how the vendor and its subcontractors may use and disclose information, including prohibiting unauthorized commercialization of it.
- Train staff, and keep human oversight and verification of the system’s outputs throughout its life cycle, not just during the pilot.
Notice what is missing from that list: accuracy scores, model names and demos. The regulators are not asking which AI you bought. They are asking who is accountable for it, what evidence exists, and what a patient can do when it gets something wrong. Those are questions a clinic can ask any vendor today, including us, and a vendor who cannot answer them in writing has told you something important.
How Align eCare is being designed against these six principles
Align eCare is in development and is not yet in clinical use, so what follows is our design, stated as design. The evidence for it has to come from implementation and a measured pilot.
- Privacy protective. Our policy commitment is that patient and clinic information will never be used to train AI models. Canadian data sovereignty is a design requirement; processing locations, backups and subprocessors still need deployment-level verification.
- Valid and reliable. The work is designed to be carried out by defined rules, including the official New Brunswick billing rules, from sources that can be shown, rather than generated freehand. Discrepancies are designed to be flagged to the physician, never silently resolved. Rules can still be incomplete, which is why testing and professional review remain necessary.
- Safe, and human rights affirming. Align eCare focuses on the paperwork after the visit: forms, referrals, prescription paperwork and billing evidence. It is not designed to diagnose or recommend treatment, and it presents reference information with its sources.
- Transparent. Every output is designed to carry its evidence with it, so a clinician, a clinic manager or an auditor can see what a document was built from.
- Accountable. Under our design, nothing is transmitted on a draft. Align eCare’s proposed eAttest™ protocol binds each document to the responsible physician’s attestation of that exact version. Verify first. Attest once. Transmit securely. Accountability also needs a doorbell: privacy questions reach a person at hello@alignecare.ca.
Our trust page and privacy white paper set this out in full, the platform page shows how the work is designed to flow, and our FAQAnswers & feature status answers the questions clinics ask us most. A measured pilot is how we intend to test these claims, in a clinic’s own numbers, before asking anyone to rely on them.
The regulators have published the standard. It is the standard Align eCare is being designed to be measured against.
Sources
- Office of the Privacy Commissioner of Canada, news release: “Canada’s federal, provincial, and territorial privacy commissioners and ombuds call for the responsible use of artificial intelligence in health and health research,” October 29, 2026. Read the release
- Resolution: Responsible use of artificial intelligence in health and health research, adopted September 15, 2026, Ottawa. Read the resolution
- Information and Privacy Commissioner of Ontario, AI Scribes: Key Considerations for the Health Sector and companion checklist, January 28, 2026.
