AI is becoming a main element of modern health care, with new technologies increasingly used for daily health activities, from interpreting clinical images and detecting early signs of disease to personalizing treatments, to assisting in surgical procedures. As these tools evolve from experimental to key transformers in the way clinical decisions are made, it becomes essential to integrate ethical and legal frameworks capable of ensuring a safe, fair, and transparent technology use, without stifling innovation.

From an ethical perspective, traditional principles of medicine remain fundamental, but their application changes when AI is involved in decision making. Autonomy requires that patients understand when AI is used in their care, and that they maintain real control over their treatment options. Beneficence and non-maleficence require AI tools to be clinically validated and continuously monitored by humans. Justice requires avoiding biased systems, by using diverse data sets, periodically auditing models, and providing sufficient transparency so that professionals can interpret and question the obtained AI results.

Three legal challenges are increasingly presented by novel AI technologies:

• Data protection. AI technologies rely on large volumes of sensitive information, such as medical records, images, laboratory results, and genetic data. Data protection must therefore be considered not only in the collection and storage of data, but also in its exchange between technological systems, its reuse for training models, and its retention throughout the AI lifecycle.
• Decision-making liability. When a diagnosis or recommendation is created by AI, it is necessary to define a traceable distribution of responsibilities among developers, institutions, and health care professionals.
• Broad regulatory observance. While regulations are rapidly evolving, the use of AI in sensitive areas such as medicine cannot be based solely on current rules; instead, it must be guided by best practices and international certifications as well, ensuring consistent compliance with the shifting regulatory landscape.

Although AI is already part of modern health care, its real value will depend on whether ethical expectations and legal requirements can be implemented to operationalize AI in daily clinical practice. Challenges in 2026 will extend beyond adopting new AI tools, to developing the institutional capacity to use them responsibly, in ways that truly enhance clinical judgment and procedures.