Hi-Tech Healthcare

AI is revolutionising healthcare: this is how medical liability is changing

From data management to defensive medicine: the implications of artificial intelligence for patient relations and the use of technology

by Claudio Testuzza

31 July 2026

Aggiungi Il Sole 24 Ore
ai preferiti su Google

3' min read

Translated by AI

Versione italiana

3' min read

Translated by AI

Versione italiana

Artificial intelligence (AI) is revolutionising the healthcare sector, driving innovation and transforming the way medical services are delivered. However, the introduction of AI-based systems also raises important issues, particularly regarding medical liability in the event of errors. Artificial intelligence (AI) in healthcare refers primarily to the application of advanced systems and algorithms capable of processing complex information to support diagnosis, treatment and patient management. To this end, AI utilises technologies such as machine learning and deep learning, which are based on neural networks that learn from data. AI algorithms are trained on large amounts of data.

Data management

The management of healthcare data using AI requires rigorous security measures. Artificial intelligence systems collect and analyse sensitive data, and a security breach could expose patients’ private information. Furthermore, it should be noted that the use of AI in healthcare may carry the risk of doctors becoming overly reliant on automated systems, with the danger of relying on assessments that are not always entirely transparent. With the increasingly rapid integration of artificial intelligence into healthcare, it is therefore becoming ever more urgent to address the question: who is responsible when AI is used to make decisions in a healthcare context?

Doctor-patient relationship

One of the main concerns is interference in the relationship between healthcare professionals and patients. This has obvious implications for professional conduct and ethics. The WHO, which had already addressed this issue in June 2021, reaffirms the importance of protecting human autonomy, confidentiality and privacy; ensuring transparency, explainability and intelligibility; and promoting responsibility and accountability.

Corporate Responsibility

One aspect that is often overlooked concerns the company’s liability when it is an individual employee — rather than the employer — who introduces artificial intelligence tools into the workplace. The prevailing legal view is that a company may be held liable for the use of AI by its employees even when it has not deliberately decided to adopt such tools.

The operator’s liability

Healthcare professionals are not exempt from liability. Their role is to verify the results generated by AI, and they are required to interpret them critically. The main legal issues concern the matter of civil liability. The use of Artificial Intelligence involves the processing of healthcare data in a complex, automated and profiling manner, which, in accordance with the GDPR, requires a solid legal basis and a broader informed consent, also aimed at permitting the use of AI in healthcare. Therefore, it is not just a question of clinical expertise, but also the ability to understand, manage and monitor the technological tools employed.

Any high-risk AI system must operate under the direct supervision of qualified healthcare staff, and every instance of its use must be recorded in the Electronic Health Record.

Manufacturer’s liability

In terms of liability, a distinction is made between AI-equipped devices under medical supervision – where the concept of professional negligence still applies – and those with a high degree of autonomy and opacity, where the breakdown of the causal link suggests that the focus should shift towards the liability of the manufacturer and the supply chain. The balance between innovation and patient protection also depends on the promotion of reliable and transparent AI, accompanied by strict liability on the part of the manufacturer in cases of ‘machine’ error.This approach gives rise to the liability of the doctor using AI, who effectively becomes a medical device operator, bearing responsibility for the choice of AI and tasked with ensuring that patients are provided with reliable tools.

Informed consent

Articles 7 and 13 of Law 132/2025 introduce the requirement for specific informed consent to the use of AI, in written or digital form, in accordance with Law 219/2017. Patients have the right to know when and how AI contributes to diagnosis or treatment. They must be informed about the function, limitations, risks and safeguards ensuring human oversight of the systems. From 10 October 2025, informed consent will no longer apply solely to the choice of treatment, but also to the AI technology involved in that treatment.

Consent thus becomes a means of informed participation and shared responsibility, strengthening the therapeutic relationship, which retains its dialogical dimension even in the digital age.

AI and defensive medicine

Whilst the introduction of AI systems in healthcare may, on the one hand, reduce defensive medicine (for example, by reducing the margin of diagnostic error), on the other hand it paves the way for a new form of digital preventive medicine, which raises ethical and legal issues and will have an impact on the allocation of civil liability. Starting from the dilemma between adherence to guidelines (Law 24/2017), good medical practice and the use of innovative but opaque tools, it is important to highlight the risk of AI-enabled treatment tools being used purely for defensive purposes.