Empowering teams with clarity, consistency, and lasting impact across sectors.

‹ Back

As I mark one year of sharing my perspectives on AI governance and ten years of consulting with ThompsonBAYTED, I’m launching into a four part editorial series entitled “From AI principles to practice: Operationalizing responsible AI through governance, leadership and continuous improvement” to explore what the innovation governance gap means for leaders, organizations and the future of responsible AI. 

Article 1 – The Policy Signal
Canada's healthcare AI moment: Innovation is accelerating. Is Governance Keeping Up?

As Canada invests in AI enabled healthcare, data infrastructure and innovation, privacy regulators call for stronger accountability and oversight, instigating a more fundamental question emerges: Do our health systems have the governance capacity to implement AI responsibly?
       Artificial intelligence is increasingly being positioned as part of the solution to some of Canada's most persistent healthcare challenges such as workforce shortages, growing demand for services, rising costs, operational inefficiencies and delays in access to care. From clinical decision support and diagnostic technologies to predictive analytics, surgical robotics and administrative automation, AI has the potential to reshape how healthcare is delivered, managed and evaluated.
       The reality is, technological capability does not automatically translate into organizational readiness. On October 29, 2026, Canada's federal, provincial and territorial privacy commissioners and ombuds issued a joint announcement calling for the responsible development, procurement and adoption of artificial intelligence in healthcare and health research. Their resolution, adopted September 15, calls on governments to strengthen legislative and regulatory frameworks and legally require relevant parties to establish effective AI governance and accountability frameworks.
       The regulators identify six essential principles: privacy protection, validity and reliability, safety, respect for human rights, transparency and accountability. This is an important development for Canada's healthcare system. It signals a growing recognition that responsible AI adoption requires more than technological investment, privacy policies or individual safeguards. It requires institutional mechanisms capable of overseeing AI throughout its lifecycle.

       The question is no longer simply whether healthcare organizations are ready to adopt AI, but whether they are equipped to govern it. Canada's AI ambitions must be matched by governance capacity, and the federal government's June 2026 AI for All strategy reinforces Canada's ambition to accelerate AI adoption, strengthen sovereign infrastructure and position healthcare as a priority sector. Among its commitments is a $100 million investment to establish a Health Sector Data Space, in partnership with the Canadian Institute for Health Information (CIHI), connecting secure, private and standardized datasets to support clinical trials, health services research and performance measurement.
       A further $100 million investment in the VITAL health data platform is intended to expand access to secure hospital data and support health research and innovation. These investments recognize that data infrastructure is foundational to AI innovation. However, infrastructure alone does not establish the conditions for responsible implementation. As healthcare organizations increasingly integrate AI into clinical and operational environments, questions of data stewardship, decision rights, human oversight, vendor accountability, transparency and performance monitoring become equally important.
       The challenge is not to slow innovation, but to ensure that innovation is supported by the organizational capacity to manage its risks, evaluate its outcomes and sustain public trust. One of the persistent challenges in responsible AI is the gap between establishing governance principles and translating them into everyday organizational practice. Many organizations can articulate commitments to ethical AI, privacy, transparency and human oversight, yet fewer have clearly defined mechanisms to demonstrate how those commitments influence procurement, implementation, operational decision making and ongoing change management and performance. This distinction matters particularly in healthcare, where AI enabled systems may influence clinical decisions, patient pathways, resource allocation and access to services. 
       Human oversight, for example, must mean more than having a person somewhere in the decision making process. It requires clearly defined authority, appropriate expertise, the ability to challenge AI generated recommendations and mechanisms to intervene when outcomes are unreliable or downright wrong. Similarly, accountability cannot end when a system is approved or a vendor agreement is signed, it must extend across the technology and the contractual life cycle. This is where governance needs to evolve from a compliance exercise into an organizational capability.
 

My research and work developing AI-RESPECT™ Responsibility, Ethics, Sovereignty, Protection, Explainability, Compliance and Transparency, explore how organizations can translate responsible AI principles into practical governance mechanisms. The framework is grounded in a broader proposition that governance capacity is itself an essential form of AI capacity. This requires organizations to move beyond static policies and one-time approvals towards an adaptive approach to governance.
       There is an important connection here to Lean management and the Plan-Do-Study-Act (PDSA) cycle commonly used in health care quality improvement initiatives. Rather than treating AI governance as a completed checklist, organizations can use continuous improvement to:

• Plan: Assess organizational readiness, intended use, data requirements, risks, decision rights and human oversight
• Do: Implement AI within defined operational parameters, with appropriate safeguards, training and accountability
• Study: Evaluate performance, reliability, equity, unintended consequences and measurable service outcomes
• Act: Adapt controls, improve processes, address identified risks or restrict and discontinue applications where necessary

This approach recognizes that AI systems, organizational conditions and operational risks evolve, and governance must evolve with them. It also creates an opportunity to connect responsible AI adoption with measurable organizational performance, rather than treating governance and innovation as competing priorities. For boards, executives and health system leaders, the emerging policy environment presents an opportunity to reconsider how AI oversight is embedded within organizational governance. Some important questions deserve greater attention:

• Does staff have a comprehensive understanding of where and how AI is being used across the organization?
• Are accountability, decision rights and human oversight clearly defined?
• Do data governance arrangements adequately address privacy, security and data sovereignty?
• Can staff independently evaluate AI performance, identify bias and respond to unintended consequences?
• Is there a measurement mechanism in place to track whether AI investments actually improve patient outcomes, access, workforce capacity and operational performance?

These are not simply technology questions, but questions of leadership, governance, organizational readiness and public accountability that all leaders must be considering.
 

Canada has an opportunity to demonstrate that technological innovation and public accountability can advance together, as recent exchanges during the UN’s Digital Cooperation Day in New York City September 21, 2026 signal that the next phases of AI all point to governance challenges, regardless of the industry. Investments are being made, policy conversations are evolving and the regulatory expectations are becoming increasingly articulated, yet the challenge is implementation.
       Healthcare organizations will need more than access to advanced technologies and increasingly sophisticated datasets, they will need the governance infrastructure, leadership capacity and continuous improvement mechanisms to ensure those investments translate into safe, equitable and sustainable improvements in care. The future of healthcare AI will not be determined solely by how advanced the technology becomes, but by how effectively our institutions govern its use.

 

In the next article of the series, I will examine Canada's emerging healthcare AI policy landscape in greater depth, including the October 29 resolution, the implications for health system leadership and the practical governance capabilities required to operationalize responsible AI. I will also explore how AI-RESPECT™ and a PDSA-based continuous improvement model can help bridge the gap between responsible AI principles and organizational accountability. Innovation creates the opportunity, but governance determines whether we can realize it responsibly.

 

CT