Artificial Intelligence in Healthcare: Effective Regulation Can Boost Use of Explainable Innovation in Healthcare

In 2023, the impact of Artificial Intelligence (AI), including Machine Learning, on our lives became evident. Now is the right time to guide use of this pure technology with a legal and ethical framework. We need to apply fair use of AI rules so that explainable AI-based technologies can also be used in clinical care and diagnostics.

AI Can Support Doctors Through Automation and Insights

Categories of Artificial Intelligence (AI), such as Machine Learning based on statistical methods, are already of great help today. I believe that AI can transform how healthcare is delivered and serve as a valuable lever to ensure that individuals have access to high-quality medical information and diagnostics. However, harnessing the full potential of AI requires solid regulation and ensuring explainability – the healthcare professional as well as the patient must know how the algorithms work. I am confident that fit-for-purpose legislation will accelerate the adoption of fair AI in healthcare, strengthening trust in such AI-powered products.

Private Sector Contributes To Setting Standards To Reinforce Patient Safety

I am an advocate for AI for the benefits highlighted before, although I am aware that some categories of AI (such as Generative AI) may not be able to guarantee explainability today and may pose some risks, opaque data gathering and processing, black box effect of neural networks, bias in the data on which large language models were trained, and AI hallucinations that are hard to track. There are still unclear elements regarding the functioning of Generative AI, so its use in medical diagnosis requires caution and care and above all confidence that fair and ethical principles can be applied. Thus fit-for-purpose legislation and international governance are needed.

AI Governance Is A Global Task, As Algorithms Cross Borders

I welcome with optimism the regulatory initiatives at the national and international level initiated in 2023 and announced for 2024. In Europe, the EU AI Act recently reached a political agreement (1). Once adopted, it will become the first legislation that creates a comprehensive set of rules, rooted in fundamental values, governing its use regardless of the type of AI technology across multiple countries.

Conclusions

AI algorithms are already safely used in healthcare to improve the precision and effectiveness of medical technology and facilitate diagnosis and treatment of patients.

A single regulation will not solve all the challenges of AI development and adoption. Yet, to navigate the complexities of this transformative technology, we need a legislative toolbox that builds upon already robust sectoral legislation e.g. EU Medical Device Regulation and In Vitro Diagnostic Medical Devices Regulation. Technology for healthcare must be not only understandable but also co-developed and covalidated by its users. The future of AI in healthcare is a matter of collaboration, transparency and fairness. Let us embrace ethical and transparent AI in healthcare and work together on solid legislation in support of AI-based applications for the benefit of patient well-being.