Vai menu di sezione

WHO – Ethics ad Governance of Artificial Intelligence for Health: WHO Guidance
Anno 2021

The World Health Organization (WHO) has developed a set of guidelines, recommendations, and directives regarding the development, application, and use of artificial intelligence (AI) technologies in the field of medicine.

The report Ethics and Governance of Artificial Intelligence for Health, published by WHO in June 2021, analyzes the main opportunities and major challenges posed by the emergence of AI in healthcare, highlighting both its benefits and risks. It recommends the development of policies, principles, and best practices for the ethical use of AI in medicine and identifies tools to prevent abuses and improper uses of AI that could undermine the protection of fundamental rights.

The document consists of nine sections and an annex.

The first section explains the reasons behind WHO’s involvement in this area and identifies the primary intended audience for the report’s guidance. Sections two and three aim to define more precisely the context in which AI is applied in medicine. Specifically, the second section provides a definition of AI, describes the various machine learning techniques used by these technologies, and defines health-related big data. The third section identifies the main applications of AI technologies in healthcare, including those used in low-income countries, such as AI in health research, drug development, health system management and planning, and public health monitoring.

The fourth section describes the laws, legal disciplines, policies, and principles that are or can be applied to the use of AI in healthcare. These include tools for protecting fundamental rights, rules governing personal data protection, and regulations and policies concerning the use of health data. This section also discusses ethical contexts relevant to AI use in medicine, and the respective roles of bioethics, law, public policy, and various regulatory frameworks as sources of ethical norms in this sector.

The fifth section outlines six key ethical principles that, according to the WHO expert group, should guide the development and use of AI in medicine. These are: protection of human autonomy; promotion of human well-being, safety, and the public interest; ensuring transparency, explainability, and intelligibility of AI systems; fostering accountability and reliability of the technologies used; ensuring inclusiveness and equity in AI development and application; and promoting sustainable AI that effectively meets human needs.

The sixth section addresses the main ethical challenges where these principles may be applied: the appropriateness of using AI technologies; AI and the digital divide; data collection and use; accountability for decisions made using AI; autonomous decision-making processes; biases and discrimination linked to AI use; AI-related safety and cybersecurity risks; the impact of AI on the labor market and healthcare employment; commercialization of AI in healthcare; and the relationship between AI and climate change.

The final sections of the report focus on identifying legal and non-legal tools and regulatory approaches that can support the ethical use of AI in medicine, suggesting appropriate governance models and offering recommendations toward that goal.

Section seven analyzes how various stakeholders can introduce good practices, programs, and ethical initiatives to meet the ethical principles and legal requirements in this domain. Possible actions include: ethically and transparently designing AI technologies; establishing mechanisms for public engagement (especially involving providers and patients) and demonstrating the reliability of the technologies used; implementing appropriate impact assessment systems; and conducting research programs on the ethical use of AI in healthcare.

Section eight addresses the issue of responsibility and how existing accountability models may evolve with the increased use of intelligent technologies in healthcare. It considers which entities might be held responsible for the selection and use of AI technologies and how liability rules may influence stakeholders' decisions. It also explores whether machine learning algorithms should be treated as products, how to compensate individuals harmed by AI technologies, and the role of regulatory agencies in this context.

The ninth section presents the essential components for establishing governance for AI in health. It examines governance models under development or already established, focusing on issues such as data governance; benefit-sharing and control; governance in the private and public sectors; regulatory and legal considerations; and the possible roles of a policy observatory, model legislation, and global AI governance.

Finally, the report offers practical guidance for implementing the WHO guidelines, targeting three groups of stakeholders: AI technology developers and programmers, health ministers, and healthcare workers. The document emphasizes that the recommendations it contains should be viewed as a starting point for further discussion and decision-making tailored to each specific context by the stakeholders involved in developing and applying artificial intelligence.

The full text of the document is available at the following link and in the download box.

Marta Fasan
Pubblicato il: Lunedì, 28 Giugno 2021 - Ultima modifica: Martedì, 01 Luglio 2025
torna all'inizio