Pan-Canadian AI for Health (AI4H) Guiding Principles

The Pan-Canadian AI for Health (AI4H) Guiding Principles outline shared values intended to guide Federal, Provincial, and Territorial (FPT) governments'Footnote 1 efforts to support the responsible and ethical adoption and use of AI technologiesFootnote 2 across Canada's health systems. As part of the Working Together to Improve Health Care for Canadians plan, Health Ministers (except Quebec) endorsed a Joint FPT Action Plan on Health Data and Digital Health and a Pan-Canadian Health Data Charter to guide the path forward for modernizing Canada's health system with standardized health data and digital tools.

AI technologies are a key component and enabler of this plan and, when harnessed responsibly, have the potential to transform health care and improve health outcomes. AI technologies also intrinsically rely on access to sufficient high-quality data that is representative of the populations' diversity, which is another reason why AI use for health care must be considered as a component of this plan.

By harnessing the transformative potential of AI in health, FPT governments will unlock unprecedented opportunities to drive research and innovation, improve health systems and advance effective, patient-centered care, ultimately leading to better health outcomes. While there are significant risks from the use of AI in health, there are also significant risks from not taking timely action to enable the broad and equitable health benefits that are possible from AI. The Guiding Principles recognize the importance of accelerating the adoption of equitable and transparent AI technologies while ensuring safeguards are in place to protect the public from potential harms.

Trust is a key enabler of AI adoption, particularly in the context of health. By adhering to these principles and through ongoing engagement with people living in Canada, including underserved populations, patients and their representatives, health care professionals, as well as innovators, researchers, developers, suppliers, purchasers, and government institutions, trust can be built and maintained across the health system. This responsibility is shared across jurisdictions and will require coordination to ensure alignment and interoperability of approaches and legal frameworks.

The following shared principlesFootnote 3 are intended to guide the collective action of FPT governments in the increased adoption of AI technologies to improve the health of all people in Canada, to modernize health systems and healthcare delivery, and promote equitable health outcomes:

  1. Person-centricity: The well-being and diverse perspectives of the people of Canada, including specific populations, are central to decisions about the adoption of AI technologies and the use of health data and health information upon which AI technologies rely including through inclusive engagement.
  2. Equity, diversity and inclusion: AI technologies are adopted in Canada's health systems in a manner that supports fair and equitable access to high-quality, culturally appropriate and safe health services to improve health outcomes while minimizing bias and reducing health inequities.
  3. Privacy and security: The use of individuals' health data in AI technologies is supported by measures to ensure health data is protected and respects individual privacy through appropriate consent and de-identification. Secure systems will enable confidentiality, integrity, availability and patient control over their personal health data, which is held by health data stewards in line with applicable privacy, information and data legislation.
  4. Safety and oversight: AI technologies are developed and used responsibly and safely in Canada's health system with multi-stakeholder oversight, including safety monitoring across the AI lifecycle to ensure the validity and reliability of AI outputs and evaluating outcomes against clear and evidence-based benchmarks. The adoption of AI for health technologies is supported by well-communicated, appropriate and effective regulatory, policy, ethical and/or procurement frameworks.
  5. Accountability and responsibility: Developers, purchasers, operators, and regulators of AI technologies used in Canada's health systems are accountable within the appropriate and ethical development, adoption, use, and oversight of AI for health to promote secure, safe and equitable health care outcomes. This includes the shared responsibility to leverage Canada's strengths and foundational investments in AI and biomedical sciences research to responsibly scale innovative technologies and deploy AI for health. Efforts to increase adoption of AI health technologies are supported by appropriate governance, accountable decision-making processes and clear guidelines developed and informed by relevant stakeholders.
  6. Transparency and understandability: The public, health care professionals, and patients and their representatives are made aware of the role that AI is playing in Canada's health systems through transparent communication of how, when, and where AI technologies, methods and tools are used to support health care decisions, and what impact they might have, including the impacts on patient care, safety, security, and privacy. Transparent communications would include parameters for the appropriate use and ownership of AI technologies, and data for the public good.
  7. AI literacy: AI literacy, including the ability to communicate, process, and understand basic health information and services, is improved through education and transparency appropriate to each group so that people and populations are empowered to make informed health decisions. Simultaneously, decision-makers, health care professionals, researchers, and innovators are equipped to use AI technologies effectively and responsibly, support effective change management and align on the use of common terminology.
  8. Robust data and data practices: Governments and health data partners continue to collaborate and share best practices to improve the quality of, secure access to, and exchange of health data that are representative of Canada's diversity, in alignment with the Pan-Canadian Health Data Charter.
  9. Indigenous-led governance and data sovereignty: All AI values and principles set out above must respect Indigenous-led governance frameworks, and support Indigenous Peoples' and communities' rights to self-determination and data sovereignty.

Footnotes

Footnote 1

Although it shares some principles of this document, Quebec intends to maintain its full autonomy regarding the planning, the organization and the management of its health system.

Return to footnote 1 referrer

Footnote 2

AI technologies refer to AI systems, which are machine-based systems that, for explicit or implicit objectives, infer, from the input they receive how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments. Different AI systems vary in their levels of autonomy and adaptiveness after deployment. (OECD, 2023)

Return to footnote 2 referrer

Footnote 3

These shared principles have been guided by existing international principles with the goal to complement and not duplicate, while taking into account Canadian realities.

Return to footnote 3 referrer

Page details

Date modified: