As we have seen over the past few years, artificial intelligence (AI) algorithms have assumed an ever-increasing role in the practice of medicine in general, and the field of medical imaging in particular. The potential power and reach of AI have led some pioneers in the field such as Geoffrey Hinton (seen right, © Reuters) to warn us about the existential threat of the technology and advocated for government intervention. He and others have expressed concerns about both the speed at which AI is developing and the direction in which it is headed.
Today’s issue of Science has published an editorial by Urs Gasser, professor of public policy and innovation at the Technical University of Munich (TUM) in Germany, in which he reports that the European Parliament has just adopted a ground-breaking draft law known as the AI Act (AIA). According to him, the AIA “aims to safeguard fundamental rights and ensure the ethical development of AI in Europe and beyond. It is the most ambitious framework to date, designed to guide the development and usage of AI.”
Although the AIA is not yet in its final form, with further refinements by the EU expected, Gasser (seen left) believes this is an opportune time for the AI research community to brace itself, with the impacts likely to reverberate beyond the borders of Europe. While some critics are concerned that the length and complexity of the document could put a damper on innovation, especially among start-up companies, proponents of the new law have argued that the promotion of AI that is trustworthy will inspire greater innovation.
If the goal of a start-up is to enter the EU market, especially in the life sciences, then familiarity with AIA requirements will be crucial. For high-risk systems there will be rigorous premarket obligations, including risk management and requirements for transparency. Gasser points out that, “a team of medical researchers deploying a customized AI in the context of personalised medicine must ensure that the training data are free from racial, social, or religious biases.”
Gasser is also optimistic the AIA will contribute positively by reshaping commercial AI research practices globally and will serve as a wake-up call for the science community by highlighting the importance of embedding AI ethics and regulation within training programmes. At TUM, where he serves as dean of the School of Social Sciences and Technology, he is committed to “such bridge-building and human-centered engineering, inspiring reform within and beyond Europe.” Perhaps Geoffrey Hinton’s fears about AI (seen above right, © Getty) will be assuaged by the leadership of scholars such as Urs Gasser.