The availability of data is a condition for the development of AI. This is no different in the context of healthcare-related AI applications. Healthcare data are required in the research, development, and follow-up phases of AI. In fact, data collection is also necessary to establish evidence of compliance with legislation. Several legislative instruments, such as the Medical Devices Regulation and the AI Act, enacted data collection obligations to establish (evidence of) the safety of medical therapies, devices, and procedures. Increasingly, such health-related data are collected in the real world from individual data subjects. The relevant legal instruments therefore explicitly mention they shall be without prejudice to other legal acts, including the GDPR. Following an introduction to real-world data, evidence, and electronic health records, this chapter considers the use of AI for healthcare from the perspective of healthcare data. It discusses the role of data custodians, especially when confronted with a request to share healthcare data, as well as the impact of concepts such as data ownership, patient autonomy, informed consent, and privacy and data protection-enhancing techniques.

In the age of Big Data and machine learning, with its ever-expanding possibilities for data mining, the question of who is entitled to control the data and benefit from insights that can be derived from them matters greatly for the shape of the future economy. Therefore, this topic should be assessed under the heading of distributive justice. There are different views on who is entitled to control data, often driven by analogies between claims to data and claims to other kinds of things that are already better understood. This chapter clarifies the value of approaching the subject of control over data in terms of (a notion of moral, rather than legal) ownership. Next, drawing on the work of seventeenth-century political theorist Hugo Grotius on the freedom of the seas, and thus on possibilities of owning the high seas, I develop an account of collective ownership of collectively generated data patterns and explore several important objections. Since control over data matters enormously and is poorly understood, we should treat questions about it as genuinely open. This is a good time to bring to bear unorthodox thinking on the matter.

Policymakers in the EU have debated whether the digital economy may benefit from the introduction of data ownership and data access rights. The chapter asks whether and how the concepts of data ownership and data access rights may serve the goal of establishing an adequate free flow of data in the digital single market.

It first maps the policy goals contained within the EU’s Digital Single Market Strategy and then analyzes how data ownership – understood as a property right – may serve the implementation of this strategy.

The conclusion is that data ownership is unlikely to further the establishment of an adequate free flow of data. Therefore, the chapter examines whether ownership, understood as control over personal data, is a viable alternative to the property rights approach.

As a final step, the question is examined if, and under what circumstances, access rights to data already exist, or should be introduced, to allow individuals and businesses to use both personal and non-personal data.