Artificial intelligence has entered a new phase in Canadian health care. BLG recently hosted health care leaders, hospital representatives, technology companies, and industry stakeholders at its Artificial Intelligence in Canadian Health Care Symposium in Toronto.
Discussions throughout the symposium revealed a clear shift in how health care organizations are thinking about AI. A year ago, many conversations focused on experimentation and use cases. As of 2026, attention is increasingly turning to implementation, governance, procurement, accountability and scale. A common theme emerged throughout the day: AI has entered a new phase in the sector.
Artificial intelligence is moving from experimentation to implementation
The conversation is no longer whether AI can be used. AI is already being tested, evaluated, and deployed across clinical, operational, and administrative settings. The more pressing question is how health care organizations can deploy, govern, and scale these tools in a way that is safe, trusted, operationally sustainable, and legally defensible.
For hospitals, health authorities, health care companies, insurers, and health technology vendors, the next 12 to 24 months are likely to be defined less by technological capability and more by governance, accountability, procurement, privacy, and implementation discipline.
AI adoption is accelerating, but scale is lagging
Health care organizations across Canada are increasingly exploring AI tools for documentation, workflow efficiency, patient flow, triage, diagnostic support, imaging review, and operational decision-making. AI is no longer confined to research environments or isolated pilots.
However, the widespread deployment of AI across health care systems remains relatively limited.
Many organizations have successfully identified promising use cases and conducted pilot projects. Far fewer have successfully integrated AI into day-to-day operations in a manner that is scalable, repeatable, and trusted by clinicians, patients, and organizational leadership.
This implementation gap is becoming one of the most significant challenges facing health care organizations. The barriers are rarely technical alone. More often, organizations struggle with governance structures, procurement models, privacy obligations, accountability frameworks, implementation planning, and organizational readiness.
The result is that health care organizations increasingly find themselves asking not whether AI works, but whether they can confidently operationalize it within the realities of the Canadian health care system.
Why this matters now
The timing of this conversation is important.
Across Canada, health care organizations continue to face significant workforce pressures, administrative burden, clinician burnout and access-to-care challenges. As a result, many organizations are evaluating AI not as a technology initiative, but as a potential operational tool to help address real constraints within the health care system. This urgency is one reason why discussions around implementation, governance and accountability have become so important.
Against that backdrop, AI is increasingly being viewed as a potential tool to reduce documentation burdens, support clinical workflows, improve operational efficiency, and assist decision-making.
Yet health care is not simply another industry adopting a new technology.
AI deployment in health care may affect patient care, clinical judgment, privacy rights, institutional accountability, and public trust. Decisions about implementation therefore carry different implications than they might in other sectors.
This is one reason why governance has become such a central theme in discussions about health care AI.
Governance is becoming a precondition to scale
A recurring theme throughout the symposium was that governance is no longer an optional consideration addressed after deployment.
Increasingly, governance is becoming the bridge between successful pilots and sustainable implementation.
Health care organizations are being required to answer a series of fundamental questions:
- Who approves the deployment of an AI system?
- What evidence is required before implementation?
- Who is accountable when issues arise?
- How should systems be monitored over time?
- When should a system be paused, retrained, modified, or discontinued?
- How should incidents, near misses, and unexpected outcomes be documented and managed?
Governance is therefore becoming less about policy development and more about establishing operational structures that support the responsible use of AI across an organization's lifecycle.
Organizations that fail to address these questions may find it difficult to move beyond isolated use cases, regardless of the underlying technology's capabilities.
Privacy and data governance remain foundational
No discussion of health care AI can occur without addressing privacy and data governance, something ever more timely now as Canada recently introduced Bill C-36, An Act to enact the Protecting Privacy and Consumer Data Act (PPCDA), which would replace the Personal Information Protection and Electronic Documents Act (PIPEDA) as our federal private-sector privacy regime; BLG published in-depth guidance on the proposed changes.
AI implementation frequently engages questions involving personal health information, de-identification, consent, vendor access, cybersecurity, secondary use of data, and re-identification risk.
As health care organizations pursue more sophisticated AI initiatives, they are increasingly being asked to balance innovation objectives against statutory obligations and public expectations.
Importantly, governance considerations extend beyond compliance. Public trust remains a critical factor in successful AI adoption. Health care organizations must be able to explain how data is being used, what safeguards are in place, and how accountability is maintained throughout deployment.
Trust is not simply a communications issue, it is increasingly becoming an operational requirement.
The standard-of-care conversation is beginning
One of the more interesting discussions throughout the symposium involved the evolving relationship between AI and the standard of care.
Historically, concerns have focused on the risks associated with overreliance on AI-generated outputs. Clinicians remain responsible for exercising professional judgment, and AI does not transfer accountability for clinical decisions.
However, health care organizations should also be mindful of a different question that may emerge as AI adoption matures.
As certain tools become more accurate, more reliable, and more widely accepted within clinical practice, questions may eventually arise regarding whether failure to consider certain widely accepted AI-supported tools creates a different category of risk.
While Canadian law has not yet provided definitive answers, organizations should anticipate increased attention to how AI intersects with professional obligations, clinical decision-making, and institutional risk management.
This issue is likely to become increasingly important as AI moves from experimentation to broader adoption.
AI procurement and contracting are becoming strategic issues
Another key takeaway from the symposium was that many AI risks are ultimately managed or created through contractual arrangements.
Health care organizations often focus on the functionality of a proposed tool. Equally important, however, are the terms governing how that tool will be deployed, maintained, updated, monitored, and supported over time.
Organizations should carefully consider:
- allocation of liability;
- privacy and security obligations;
- data ownership and permitted uses;
- implementation responsibilities;
- performance expectations;
- audit and reporting rights;
- model updates, functionality and retraining obligations;
- service interruptions and incident response; and
- governance and monitoring requirements.
As AI systems continuously evolve, procurement arrangements increasingly need to address lifecycle management rather than simply software acquisition.
This makes contracting a strategic component of AI governance, not merely an administrative exercise.
Five practical questions for health care organizations
As health care organizations evaluate AI initiatives, leadership teams should consider:
- What problem are we trying to solve?
- What governance structure will oversee implementation and ongoing use?
- How will privacy, data governance, and cybersecurity risks be addressed?
- What contractual protections are required to manage risk appropriately?
- How will the organization evaluate performance, monitor outcomes, and respond to issues after deployment?
Organizations that can answer these questions effectively will be positioned to move from experimentation toward sustainable implementation.
Why health care AI requires a multidisciplinary approach
The issues associated with health care AI do not fit neatly within a single legal or operational discipline.
Few health care AI initiatives involve a single issue or practice area. Questions involving AI frequently engage health law, privacy, cybersecurity, technology procurement, regulatory compliance, governance, insurance, litigation risk and enterprise risk management simultaneously.
As a result, organizations increasingly require advice that integrates multiple perspectives rather than addressing individual issues in isolation.
At BLG, our Health Care, Cybersecurity & Privacy, Technology, Corporate Commercial, Regulatory, and Litigation teams work together to help health care organizations assess AI initiatives from multiple angles, whether evaluating governance frameworks, negotiating procurement arrangements, assessing privacy obligations, developing implementation policies, or managing operational and clinical risk.
Looking ahead
If there was one overarching takeaway from BLG's Artificial Intelligence in Canadian Health care Symposium, it was this: AI has moved from innovation to implementation.
The organizations that succeed over the next 12 to 24 months will not necessarily be those that adopt AI first. They will be the organizations that can govern it, contract for it, explain it, monitor it, and stand behind it once it becomes part of day-to-day operations.
As health care organizations move from experimentation to operational deployment, the ability to govern, monitor and manage AI will increasingly distinguish successful initiatives from unsuccessful ones. BLG will continue working with hospitals, health care organizations, technology companies and other sector participants as they navigate the legal, regulatory, governance and operational challenges that accompany this next phase of AI adoption.
Reach out to any of the authors or key contacts below to discuss how AI governance, procurement, privacy or clinical risk considerations may affect your organization.
Please also sign up for BLG’s AI x Health Care newsletter that aims to provide periodic updates on the latest AI in health care news.
