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

Data privacy issues The amount and types of data generated today in healthcare settings significantly outpace the ability of humans to consume, comprehend, and use to inform non-trivial patient care decisions compared to AI tools, which can process these information and come up with a prompt decision. The AI tools applied in healthcare settings require vast amounts of high-quality data to learn to perform efficiently and to perform their functions properly. However, this reliance on high-quality data creates enormous concerns regarding data privacy. 474 These concerns, especially in healthcare settings, are warranted, where the data required for training AI systems and their successful functioning involves sensitive and confidential patient information. In the event of incidents of data leakage or misuse, there are serious consequences, as it could result in serious harm to patients and healthcare providers since this data predominantly contains sensitive patient information, including confidential conversations between healthcare professionals and their patients, patients’ health records, and identity information. 475 In today’s world, the capacity of individuals to manage how personal data is kept, updated, and shared between parties is critical to data privacy. Recently, with the introduction of powerful internet-based data mining tools, data privacy-related issues have become rampant, making data privacy and control over personal information increasingly crucial. 476 Individuals, while applying AI tools to promote their access to healthcare, have limited oversight over what passive data is collected and how that data is transformed into a recommendation or healthcare decision, limiting their ability to challenge any decisions made and may result in a loss of personal autonomy, as well as raise data privacy issues. 477 The critical components of privacy protection and AI applications to address access to healthcare issues add to the risks to individual privacy. This risk is exacerbated, particularly in developing countries, which are unlike many developed nations that have strong data protection laws that aim to protect the privacy of their citizens, including countries like the United States, which passed the Health Insurance Portability and Accountability Act (HIPAA) in 1996 to strengthen the law to protect healthcare facilities, and the General Data Protection Regulation (GDPR) applicable in the European Union which acts as the data protection and privacy regulation that prescribes stringent rules that must be respected while working with data belonging to individuals. The risks involved in applying AI tools to address access to healthcare issues in developing nations include pertinent risks that must be prevented by the adoption 474 Davenport T, Kalakota R, ‘The Potential for Artificial Intelligence in Healthcare’ (2019) 6 Future Healthcare Journal 94. 475 Kitkowska A, Karegar F and Wästlund E, ‘Share or Protect: Understanding the Interplay of Trust, Privacy Concerns, and Data Sharing Purposes in Health and Well-Being Apps’ (2023) CHItaly 2023: 15th Biannual Conference of the Italian SIGCHI Chapter . Copeland BJ, ‘Artificial Intelligence’, Encyclopaedia Britannica accessed 15 October 2023. 476 Manheim K and Kaplan L, ‘Artificial Intelligence: Risks to Privacy and Democracy’ (2019) 21 Yale JL & Tech 106. 477 Kleinpeter E, ‘Four Ethical Issues of E-Health’ (2017) 38 IRBM 245.

113