Community hospitals building a regional AI governance collaborative

Artificial intelligence has quickly moved from the periphery of healthcare innovation to the centre of clinical, operational and administrative transformation. Ambient clinical documentation, predictive analytics, imaging support, scheduling optimization and administrative automation are no longer future possibilities – they are today’s procurement decisions.

While AI adoption is accelerating, governance has not kept pace.

Across Canada, hospitals are independently evaluating the same AI vendors, conducting duplicate privacy and cybersecurity assessments, creating separate policies, and asking identical ethical questions.

This fragmented approach consumes scarce resources, produces inconsistent standards and leaves smaller organizations struggling to access the expertise required to evaluate increasingly sophisticated technologies.

Recognizing this challenge, a group of hospitals across Ontario’s Champlain region are coming together to build a Regional AI Governance Collaborative – an innovative model to enable responsible AI adoption through shared expertise, common governance principles and collective evaluation.

This initiative is co-lead by Lindsay Wyers, VP, digital transformation and CIO from Queensway Carleton Hospital, and Scott Coombes, VP and CFO from Pembroke Regional Hospital.

The goal is simple: collaborate on governance so hospitals can remain autonomous in implementation.

The problem with everyone working alone: Healthcare organizations face growing expectations from clinicians, patients and regulators to ensure AI systems are safe, equitable, transparent and trustworthy. Every new AI solution requires review from multiple disciplines including privacy, cybersecurity, legal, clinical operations, ethics and information technology.

For many organizations, particularly medium-sized and community hospitals, assembling this expertise for every procurement is becoming increasingly difficult.

Without collaboration, organizations often duplicate the same work:

•  Privacy impact assessments
• Threat risk assessments
• Vendor due diligence
• Clinical safety reviews
• Policy development
• Ethical assessments
• Procurement evaluations

The result is rising governance costs, inconsistent standards across organizations and slower implementation of technologies that could improve patient care.

Rather than asking every hospital to build identical governance capabilities independently, the Collaborative asks a different question: What governance work can be done once and shared many times?

A new model for shared governance: Unlike a traditional committee or community of practice, the Regional AI Governance Collaborative has a clearly defined mandate.

The Collaborative will not decide which AI products individual hospitals must implement. Nor will it become an approval body or replace organizational decision-making.

Instead, it will develop the foundational governance infrastructure that every organization requires.

This includes shared AI principles, governance frameworks, risk classification methodologies, evaluation templates, procurement guidance and reusable documentation.

Member organizations will bring forward AI products or initiatives for evaluation, allowing expert working groups to perform structured assessments that can be leveraged across participating hospitals.

Each organization retains full authority over procurement, implementation and operational oversight.

This distinction is critical.

Regional collaboration creates consistency where consistency matters, while preserving organizational autonomy where local context matters most.

Building trust through expertise: One of the first questions organizations ask is straightforward: Who should evaluate AI?

The Collaborative recognizes that trustworthy AI cannot be assessed from a single perspective.

Evaluations require multidisciplinary expertise spanning clinical practice, privacy, cybersecurity, legal, ethics, procurement, information technology and operational leadership.

To support the development of a practical and sustainable governance model, the Collaborative has engaged external expertise from Info-Tech Research Group. Through the guidance and support of Krizia Francisco and Justin St-Maurice, the initiative is leveraging evidence-informed approaches, industry experience and structured methodologies to help shape AI governance practices that are scalable for healthcare organizations.

The Collaborative is also partnering with academic expertise to strengthen the ethical foundations of AI decision-making. Through collaboration with the University of Ottawa, PhD student Amanda Maria Kutenski will support the development of an ethical AI framework and decision-making tool.

This work will help organizations evaluate AI solutions through critical considerations including transparency, accountability, fairness, privacy, safety, human oversight and patient impact.

Together, these partnerships ensure the Collaborative is grounded not only in operational realities but also in emerging best practices for responsible AI adoption.

Rather than relying on individual opinions, assessments will be conducted using standardized evaluation frameworks grounded in eight foundational principles co-developed by participating organizations.

These principles emphasize accountability, patient safety, fairness, transparency, human oversight, technical robustness, patient autonomy and fiscal responsibility.

Collectively, they establish consistent expectations regardless of which organization is considering an AI solution.

The Collaborative is also exploring skills-based participation, ensuring evaluations are completed by individuals with relevant expertise rather than organizational title alone. Rotating representation will broaden participation while maintaining consistency in decision-making.

Moving beyond principles to deliverables: The Collaborative has intentionally focused on producing reusable assets rather than discussion papers. Planned deliverables include:

•  Standardized AI governance policies
• Risk assessment methodologies
• Vendor evaluation scorecards
• AI intake and review processes
• Shared procurement guidance
• AI inventories
• Governance documentation
• A curated catalogue of evaluated AI solutions
• Ethical AI assessment tools to support responsible decision-making

Hospitals will be able to leverage these resources immediately, dramatically reducing duplicated effort while improving the quality and consistency of governance decisions.

Rather than beginning every AI evaluation from scratch, organizations will have access to structured, evidence-informed assessments that can be adapted to local needs.

Early progress: The initiative has already brought together more than 20 leaders from across participating organizations, representing physicians, patients, privacy, finance, information technology, clinical operations and executive leadership.

Through a structured series of collaborative design sessions, participants have already developed:

•  Eight foundational AI principles
• Draft governance boundaries
• A proposed governance model
• Initial charter concepts
• Regional governance roadmap
• Shared understanding of roles and responsibilities

The Collaborative has also established partnerships with external experts and academic contributors to strengthen the framework and ensure it reflects both healthcare realities and emerging expectations for ethical AI. Importantly, the group has clearly defined what belongs at the regional level versus what remains the responsibility of individual organizations.

For example, while the Collaborative will develop governance frameworks and evaluation methodologies, hospitals will continue to own implementation decisions, operational monitoring and procurement authority.

A roadmap that grows with AI: Rather than attempting to govern every AI application immediately, the Collaborative is taking an iterative approach.

The initial phase focuses on establishing governance structures, completing an inventory of AI currently deployed across member organizations and applying standardized risk assessments to the highest-priority tools.

Future phases will introduce formal AI intake processes, governance workflows and monitoring approaches for higher-risk technologies, while continuously adapting governance practices as AI capabilities evolve. This recognizes an important reality: AI governance is not a one-time project but an ongoing capability.

Creating value beyond the region: Although the Collaborative is being established within the Champlain region, its potential extends well beyond participating organizations.

The governance frameworks, evaluation methodologies and reusable documentation being developed are intentionally designed to be scalable. As hospitals across Canada face similar challenges, shared governance models may offer an alternative to each organization independently building AI oversight capabilities.

Regional collaboration also strengthens the collective voice of healthcare organizations when engaging vendors, encouraging greater transparency around privacy, cybersecurity, clinical evidence and model performance.