AI Co-Pilots Set to Disrupt the Great Medical Record Oligopoly 

From: Will Falk and Keldon Bester 

To: Canadian Healthcare Observers 

Date: September 22, 2026 

Re: AI Co-Pilots Set to Disrupt the Great Medical Record Oligopoly

AI now has a credible path to disrupt Canada’s electronic medical record (EMR) markets. Not through wholesale replacement, but by working above and beside existing systems. 

Ambient scribes and second-screen clinical decision support (CDS) tools are already changing the lived experience of care without ripping out core infrastructure. 

Ambient scribes are used by more than 25 percent of Canadian primary care physicians and adoption is accelerating in both ambulatory and hospital settings. Several hospitals are deploying rapidly by relying on OntarioMD and Infoway procurement agreements as safe harbours. 

About 20 vendors have been vetted for security, privacy, liability coverage, and enterprise readiness. Major international players not on those lists, including Microsoft Dragon and Heidi, have entered anyway. 

Concurrently, best-in-class CDS tools have arrived through a different route. Products like OpenEvidence and Doximity operate without personal health information, live on a second screen, and don’t integrate with EMRs. 

That no-personal-information, no-EMR design choice has enabled fast adoption. Doximity’s acquisition of Montreal-based Pathway.MD for US$63 million shows that Canadian technology is part of this story. Vendors claim that more than 40 percent of US physicians are regular users. 

Canadian usage data are unavailable; anecdotally they are already widespread. They are a safer and more reliable substitutes for foundation AI models or the Internet. 

Late 2025 moved quickly. First, the mainstream EMR vendors validated the model and a host of providers now combine ambient documentation, second-screen CDS, and a growing list of functions. 

They can prepare patient summaries before visits, support patient communication, coach clinicians during encounters, provide translation services, draft notes, referrals, prescriptions, and sick notes, and verify clinical decisions against current evidence. 

Crucially, they do this without needing to replace the EMR. They intermediate information flow, reduce documentation burden, and offload cognitive work. The EMR remains the system of record. The co-pilot becomes the system of work. 

In Ontario, several hospitals have now stated explicitly that any scribe on the OntarioMD list may be used immediately in hospital settings without further local approvals. 

That effectively allows co-pilots to operate across both acute and ambulatory EMRs. For the first time, a single clinical layer can span systems that have historically been kept apart. 

This is disruptive given Canada’s EMR market structure. Both acute and ambulatory markets are three-firm oligopolies. In acute care, Epic, Oracle Health, and Meditech dominate hospital implementations. 

In ambulatory care, TELUS Health, WELL, and Accuro (owned by Shoppers Drug Mart/Loblaw) dominate. The result looks familiar to anyone with a Canadian cellphone plan: Three firms, roughly 80-percent market share, high switching costs, and limited competition. 

The Competition Bureau’s 2022 digital health market study documented structural barriers to competition. These include restricted data access that functions as de facto data blocking, procurement frameworks that entrench incumbents, and high switching costs. 

Health Canada and its agencies have raised similar concerns. Bill C-72 attempted to address these issues but died on the order paper. More recently, the Competition Bureau obtained a court order to compel records in its investigation of WELL Health’s acquisition of HEALWELL and Orion Health. 

The Bureau’s 2024 Competition in the AI Era summit signaled that generative AI changes the competitive calculus across sectors. The durability of EMR oligopolies has rested on friction. Workflow integration has been framed as synonymous with EMR integration. One vendor, one system, one tightly controlled environment. 

Third-party innovation was tolerated at the margins and constrained through contractual, technical, and commercial barriers. Many startup executives have complained privately about anti-competitive behaviour. A few have spoken publicly. Most stay quiet, fearing retaliation. 

A clinician co-pilot breaks that logic. If a physician can move from a clinic running Accuro to its inpatient site on Cerner and rely on the same co-pilot in both settings, a new competitive layer emerges. The co-pilot follows the clinician, not the institution. The EMRs’ moat begins to drain. 

That shift has several implications. 

First, it reframes workflow integration. Integration isn’t just embedding everything in the EMR. Second, it lowers entry barriers for third-party innovation. A co-pilot can compete on usability, intelligence, and trust without displacing the underlying system. Third, it destabilizes the neat separation between acute and ambulatory markets. Vendors may be able to compete across settings. 

Incumbents are likely to respond. When threatened, EMR vendors historically restrict access. With AI co-pilots, this may appear as limits on data read-write permissions unless formal commercial relationships are in place. Patient safety rhetoric will be deployed selectively to justify blocking data flows that would otherwise benefit patients. Acquisition and suppression of competitors will remain an attractive strategy. 

Regulators should anticipate this behaviour and act early, using data portability requirements, interoperability enforcement, and merger review powers to protect nascent competition. New AI features should be loosely coupled. 

There is reason for optimism, but only if regulators intervene quickly and deliberately. AI co-pilots offer a credible chance to reset competitive dynamics, reduce clinician burden, and improve patient care. 

Will Falk is a retired management consulting partner and Senior Fellow at the C.D. Howe Institute. Keldon Bester is executive director of the Canadian Anti Monopoly Project (CAMP). 

To send a comment or leave feedback, click here. 

The views expressed here are those of the authors. The C.D. Howe Institute does not take corporate positions on policy matters. 

A version of this Memo first appeared in Canadian Healthcare Technology magazine.