Getting Ready for the EU AI Act in Healthcare. A call for Sustainable AI Development and Deployment

Healthcare AI systems must prepare now for EU AI Act compliance: The EU AI Act came into force in August 2024, with key provisions taking effect in August 2026. Healthcare organizations and developers should begin proactive steps immediately to ensure their AI systems—whether currently deployed, in development, or planned—will meet legal requirements, rather than waiting until the deadline.

Trustworthiness assessments are essential, not optional: The paper emphasizes that evaluating trustworthiness has become central to responsible AI development in healthcare. This means combining technical performance, evidence-based validation, and ethical considerations into a comprehensive assessment framework, not treating compliance as a "formalistic exercise."

Ethical principles should guide compliance, not just legal checklists: Rather than viewing AI Act compliance as a box-checking exercise, the authors argue that developers and deployers should commit to the ethical principles underlying the Act (such as fairness, transparency, and accountability). These principles help interpret the Act's requirements and identify good practices that make AI systems more sustainable and trustworthy over time.

Alignment of technical, ethical, and legal practices is urgent in high-stakes fields: Healthcare is identified as a particularly high-stakes domain where AI decisions directly affect human health and safety. This makes it especially important to ensure that technical development, ethical standards, and legal compliance work together rather than in isolation.