New Technologies in International Law / Tymofeyeva, Crhák et al.

Following this examination of the potential to promote the right to health through the adoption of AI, on the one hand, AI holds the promise of being a catalyst in accelerating access to healthcare and allowing developing nations to leapfrog over some traditional obstacles that affect access to healthcare. However, on the other hand, this adoption of AI brings along challenges that exist in the form of technological, ethical, and legal challenges, which must be addressed to ensure proper promotion of the right to health rather than perpetuating further harm by limiting access to health for specific individuals. Addressing these issues is vital because the application of AI in public health systems, particularly to data gathering, diagnosis, and interpretation of medical data, raises significant concerns, mainly due to the sensitive and confidential nature of healthcare delivery. This especially borders on issues of trust and reliability, which become prominent as AI-driven healthcare systems carry sensitive health information and high-end patient vulnerabilities. 464 The subsequent sections of this paper shall address these issues in detail and postulate solutions for creating strong and effective legal and regulatory frameworks, which will guarantee a proper implementation of AI systems to address access to healthcare issues in the public healthcare sector of developing nations. 2. Challenges encountered applying AI to right to health issues Patient safety Patient safety is a primary challenge when discussing AI’s application to solving access to health issues today. The black-box design of most AI systems (which means that these AI algorithms typically fail to explain why a given input data produces a corresponding output) makes it challenging to provide a reason for the AI system’s decisions or to portray its findings logically or reasonably. 465 This issue mandates AI algorithms to be transparent, interpretable, and explainable to retain human agency and patient autonomy concerning treatment decisions, which is especially important for promoting the right to health and access to healthcare because the right to health is based on principles of safeguarding human dignity and promoting the well-being of people. 466 These principles are vital and recognized in several policy guidelines on AI development and relate to a critical implication when balancing ethical implications of patient safety and autonomy. 467 The black-box model of AI systems creates a problem for healthcare practitioners who apply AI to patient care and makes it impossible to inspect the decision-making process of the AI system, leading to a lack of understanding of the AI’s decisions. 468 This 464 Joshi S et al, ‘Modeling Conceptual Framework for Implementing Barriers of AI in Public Healthcare for Improving Operational Excellence: Experiences from Developing Countries’ (2022) 14 Sustainability 11698. 465 Tonekaboni S, Joshi S, McCradden MD, Goldenberg A, ‘What clinicians want: Contextualizing explainable machine learning for clinical end use’ (2019) 1 Proceedings of Machine Learning Research 21. 466 Floridi L and Cowls J, ‘A Unified Framework of Five Principles for AI in Society’ (2019) 1(1) Harvard Data Science Review 1. 467 Gregory A and Halff G, ‘The Damage Done by Big Data-Driven Public Relations’ (2020) 46 Public Relations Review 101902. 468 Kim H and Xie B, ‘Health Literacy in the eHealth Era: A Systematic Review of the Literature’ (2017) 100 Patient Education and Counseling 1073.

111