Artificial intelligence’s introduction into medical practice is rapidly proceeding and it will have important implications for patient autonomy. Intelligent prognostic, diagnostic and treatment tools will reconfigure patient choices and enable different kinds of human-computer interactions. It may also undermine patient understanding of their clinical situation and treatment options and, perhaps most significantly, the design of the technology may introduce covert evaluative elements into individual clinical decisions.

Currently it is still an open question how the law will respond to these challenges. From the situation in the United States and the United Kingdom, where there has been a noticeable push to implement medical artificial intelligence, one can see that bespoke legislative or common law solutions to these issues are unlikely to be forthcoming. Rather, it is probable that the law will shape the use of this technology by applying pre-established values, principles and rules.

This presentation takes a forward-looking approach and investigates how such an application of legal norms may proceed. It begins by highlighting features and examples of clinical artificial intelligence and outlines the general autonomy-related challenges that these pose. Next it discusses how these challenges can be located in the law of the two aforementioned jurisdictions by focussing on the role that the bioethical concept of autonomy plays in their legal orders. Finally the presentation provides a specific example of how this legally-anchored concept can then be used in the application of specific legal instruments to the novel situations involving artificial intelligence.