Chassang Gauthier, Béranger Jérôme, Rial-Sebbag Emmanuelle
CERPOP, Université de Toulouse, Inserm, UPS, 31000 Toulouse, France.
Genotoul Societal Platform, Ethics and Biosciences, GIS Genotoul Occitanie, 31000 Toulouse, France.
Int J Environ Res Public Health. 2025 Apr 4;22(4):568. doi: 10.3390/ijerph22040568.
This paper discusses the responsible use of artificial intelligence (AI) in public health and in medicine, and questions the development of AI ethics in international guidelines from a public health perspective. How can a global ethics approach help conceive responsible AI development and use for improving public health? By analysing key international guidelines in AI ethics (UNESCO, WHO, European High-Level Expert Group on AI) and the available literature, this paper advocates conceiving proper ethical and legal frameworks and implementation tools for AI in public health, based on a pragmatic risk-based approach. It highlights how ethical AI principles meet public health objectives and focuses on their value by addressing the meaning of human-centred innovations, transparency, accountability, diversity, equity, privacy protection, technical robustness, environmental protection, and post-marketing surveillance. It concludes that AI technology can reconcile individual and collective ethical approaches to public health, but requires specific legal frameworks and interdisciplinary efforts. Prospects include the development of supporting data infrastructures, of stakeholders' involvement to ensure long-term commitment and trust, of the public's and users' education, and of international organisations' capacity to coordinate and monitor AI developments. It formulates a proposal to reflect on an integrated transparent public health functionality in digital applications processing data.
本文探讨了人工智能（AI）在公共卫生和医学中的合理使用，并从公共卫生角度对国际准则中人工智能伦理的发展提出质疑。全球伦理方法如何有助于构思负责任的人工智能开发和使用以改善公共卫生？通过分析人工智能伦理方面的关键国际准则（联合国教科文组织、世界卫生组织、欧洲人工智能高级专家组）和现有文献，本文主张基于务实的风险方法，为公共卫生领域的人工智能构思适当的伦理和法律框架以及实施工具。它强调了符合伦理的人工智能原则如何实现公共卫生目标，并通过探讨以人为本的创新、透明度、问责制、多样性、公平性、隐私保护、技术稳健性、环境保护和上市后监测的意义来关注其价值。结论是，人工智能技术可以协调公共卫生的个人和集体伦理方法，但需要特定的法律框架和跨学科努力。前景包括开发支持性数据基础设施、让利益相关者参与以确保长期承诺和信任、开展公众和用户教育以及提高国际组织协调和监测人工智能发展的能力。它提出了一项建议，以思考在处理数据的数字应用中纳入透明的公共卫生功能。
