WHO - Guidance on Ethics and Governance of Artificial Intelligence

Abstract

Background: Artificial intelligence opens up a world of possibilities for public health and medicine. Algorithms "trained" on data take on automated tasks, without the need for manual programming. The WHO recognizes this potential, but also warns about the ethical challenges that AI imposes on Health Systems, professionals and patients. Problem: Artificial Intelligence can revolutionize health, but it is necessary to verify to what extent its implementation should consider the following challenges: algorithmic bias; data privacy and security; transparency and explainability; digital inequality; high costs; integration with existing systems and adequacy of legislation. Objectives: to present problems and issues related to AI, with an assessment of the impact on society, as well as to present possible guidelines to mitigate the risks of AI. Methods: based on WHO recommendations, a critical-narrative review was carried out. Results: The WHO proposes a set of ethical principles to guide the use of AI in health. The goal is to ensure that this technology is used responsibly, fairly, and transparently, benefiting everyone. The ability of AI to benefit patients and communities depends on a concerted effort not only in its implementation, but also in the creation of standards that ensure that the development of AI technologies respects human rights. AI has the power to revolutionize healthcare, from improving patient care to combating pandemics, but it requires preparation and knowledge. Investing in education and training for healthcare professionals is essential. Healthcare systems need to be adapted to integrate AI safely and efficiently. AI can empower patients and communities to take control of their own health, better understand their changing needs, effectively detect health risks, and keep people at the center of decisions about their health. Conclusions: Ethical principles for AI in Healthcare aim to lay the foundation for a responsible future, considering the well-being of all people, without discrimination or bias. It is necessary to ensure fairness and equity in accessibility, regardless of income, location, or social status; transparency and explainability for individuals and communities of decisions made by AI; ensuring the accountability of developers, users and governments regarding the consequences of the use of AI; privacy and data security must always be protected; it is necessary to promote the participation and inclusion of civil society and stakeholders, giving them a voice in the definition and use of AI in health.