The algorithm will see you now: ethical, legal and social implications of adopting machine learning systems for diagnosis and screening.

Diagnosis and screening are central in health systems. Diagnosis is key to clinicians’ professional identity, makes clinical pathways available to patients, and carries substantial medicolegal risk. Authority and responsibility to diagnose conditions and interpret test results has, until recently, belonged uniquely to clinicians. But this is rapidly changing. Machine learning (ML) systems—algorithms or forms of artificial intelligence (AI) that can learn from large quantities of data—are increasingly able to contribute to, or even take over, diagnostic and screening tasks. The big idea driving our application is this: algorithmic diagnosis and screening are imminent, and health services are not ready. We need to prepare now for the ethical, legal and social implications (ELSI) of diagnostic and screening AI. Our team includes ethicists, social scientists, lawyers, clinicians, public health academics, health economists and data scientists. We are answering international calls for a new kind of AI ELSI research based on concrete cases and engaged deeply with frontline stakeholders. We will focus on two cases where ML is well-developed: breast cancer and cardiovascular disease. We will work with clinicians, regulators, professional organisations, policy makers, data scientists, patients and the general public, to discover how ML is developing and how informants think it should be developing in these cases. Based on this we will develop strategies to address the ELSI of diagnostic and screening ML, for our cases and more generally. Our multidisciplinary, practice- and public-engaged, multi-method approach is highly innovative; we offer an independent voice in a highly privatised and proprietary field. This project is the first of its kind in the world: it will locate Australia as a leader in considering the profound ethical, legal and social implications of allowing or prohibiting artificial intelligence systems that can diagnose patients and screen people for disease.