Inside Canada’s Largest AI Scribe Evaluation: What the Evidence Shows

^{This article is an independent analysis of publicly available evaluation data.}

Canada Health Infoway’s AI Scribe Program represents a turning point for clinical documentation in Canada. Not because AI scribes are new, but because this is the first time they are being evaluated at national scale, across provinces, practice models, and real-world clinical conditions.

Launched in June 2025, the program aims to provide funded access to AI scribes for 20% of Canada’s primary care providers during 2025–26. As of September 10, 2025, 10,554 primary care clinicians had already enrolled.

This matters because it moves AI scribes out of small pilots and into everyday clinical practice, where variation, complexity, and sustainability determine whether tools actually deliver value.

Why Canada Health Infoway launched the AI Scribe Program

The motivation behind the program is clear when you look at the baseline data.

Before implementation, Infoway surveyed 2,658 primary care providers across Canada. The results paint a stark picture of administrative burden and burnout:

• 83.3% said their administrative workload was tiring
• 89% reported symptoms of stress or burnout
• 89.6% said admin burden negatively affected their wellbeing or job performance

On average, clinicians reported spending:

• 38.9 hours per week on direct patient care
• 10 hours per week on administrative work during clinic hours
• 7.2 hours per week on administrative work after hours

Nearly half (45.2%) said they spent a high or excessive amount of time on EHR work at home.

Just as importantly, clinicians overwhelmingly linked documentation burden to patient experience:

• 92.1% said administrative tasks reduce time available for patient care
• 81.8% said detailed note-taking compromises patient engagement
• 71.8% said admin tasks reduce the quality of interactions

This is not an individual productivity problem. It’s a system-level constraint on capacity, presence, and sustainability.

What makes this evaluation different

Unlike many early AI scribe pilots, Infoway’s program pairs scale with independent evaluation.

The interim evaluation is being conducted by the Women’s Institute for Health System Improvement (WIHV) and uses a mixed-methods approach that includes:

• Surveys
• Clinician interviews
• Chart reviews
• Utilisation and encounter data

That combination matters. It allows Infoway to look beyond whether AI scribes can save time in theory and assess what actually changes in day-to-day clinical work, including experience, data quality, and cognitive load.

Usage data reinforces this real-world focus. In August 2025 alone:

• 49% of enrolled clinicians were active users (defined as 10 or more uses per month)
• 708,654 clinical encounters were captured using AI scribes

This is not hypothetical adoption. It reflects routine use in busy practices.

Early signals that matter

Early interview findings help explain what those usage numbers mean in practice. While early results do not tell the full story, they do indicate where AI scribes are already making a measurable difference.

Among the first 22 clinician interviews completed as of September 2025:

• 95.4% said AI scribes reduced their administrative burden
• 86.4% reported meaningful time savings
• 45.5% said they could complete documentation during the workday rather than after hours
• 90.9% reported reduced cognitive or mental load
• 90.1% said they felt more present with patients

Presence stands out as a critical outcome. When clinicians are not mentally tracking what needs to be documented next, they can focus fully on the patient in front of them. That shift affects trust, communication, and care quality.

At the same time, the evaluation is clear-eyed about what remains unresolved. Clinicians continue to:

• Review and edit AI-generated notes for accuracy and completeness
• Encounter challenges in complex scenarios such as physical exams, multi-speaker visits, and nuanced clinical reasoning
• Carry administrative burden beyond the narrative note, including forms, inbox work, and structured documentation

These findings reinforce an important point: AI scribes can meaningfully reduce burden, but outcomes depend on workflow fit, note quality, and what happens after the visit.

What this tells healthcare leaders

The evidence does not suggest that AI scribes are a silver bullet. It suggests something more practical and more useful.

AI scribes work when they are implemented thoughtfully.

The evaluation highlights several factors that consistently shape outcomes:

• How well the tool fits into existing workflows
• Whether clinicians trust the quality and structure of the notes
• How much cognitive effort remains during and after the visit
• The level of onboarding and support provided as teams adopt the tool

These findings mirror what has surfaced across Ontario evaluations and community-based case studies: documentation tools deliver the most value when they reduce cognitive load, fit real workflows, and support the work that happens beyond the note itself.

Where Tali fits in this national evaluation

As an approved vendor in the Canada Health Infoway AI Scribe Program, Tali is being used by clinicians across a range of primary care settings, contributing real-world usage and experience data to the national evaluation.

The themes emerging from the interim evaluation align closely with how Tali is designed and implemented in practice. Reducing documentation burden is not just about transcription speed. It depends on:

• Producing notes clinicians can trust with minimal editing
• Supporting workflows beyond the narrative note
• Reducing cognitive load during the visit
• Providing onboarding and ongoing support that reflects real clinical work

As the Infoway evaluation continues, later phases will provide deeper insight into longitudinal use, note quality over time, and how administrative burden shifts across different practice models. For healthcare leaders, the opportunity is not just to adopt AI scribes, but to apply this evidence to implementation decisions that support clinicians over the long term. 

Canada now has early data on what works. The next step is applying it thoughtfully, at scale, and with clinicians at the centre.

Got questions?

Our team is here for you. 

Contact us at: help@tali.ai

Or visit : community.tali.ai

🔗 Start Free Trial

🔗 Talk to Sales