Artificial Intelligence (AI) represents a paradigm shift in society in general and in healthcare in particular. Why? Because never before has a technology had the potential to transform society as we know it.

AI poses an enormous challenge to society, including the healthcare sector. The pillars of pluralistic democracies could be profoundly and decisively affected by the generalization of AI if it is implemented without effective and systematic human oversight. Freedom of expression, the rule of law, the labor market, and the justice system are already affected by new digital technologies, and AI has the potential to expand their applications but also to increase the risks of their misuse, especially with regard to the protection of individual privacy and the processing of personal data.¹

AI has several characteristics that place it at a fundamentally different level from any other technology. In particular, the development of logical intelligence that mimics human reasoning and the absolute mastery of human language mean that the processes used cannot be explained, and therefore humans cannot understand the steps that led to a given conclusion.²

In health, everything will be different with AI. From scientific research, which is also reinventing itself in the face of its own paradigm shift, as advocated decades ago by Thomas Kuhn (1970), to medical education and other health professions, the evolution is profound and irreversible.^3,4) In particular, the advent of generative AI (of which ChatGPT is one of the most striking examples) is challenging axioms that were thought to be immutable, such as authorship of a scientific article, the redefinition of the concept of plagiarism, the exercise of patent rights for new biomedical technologies (if they are indeed invented by generative AI), among others, forcing a rethinking of the concept of scientific integrity.

Clinical practice will also face unprecedented technical and ethical challenges that will affect patient and family interviews, differential diagnosis, ancillary diagnostic tools used, proposed treatments, surgery performed by fully autonomous robots, and even long-term and palliative care.

But it is in the realm of the clinical relationship with the patient, especially the doctor-patient relationship, that the impact of AI will be most felt.⁵ This relationship must always be authentically human, and this will remain the key differentiator for the widespread use of AI in medicine and healthcare. Empathy, compassion, and dedication to patients will always be attributes of healthcare professionals that should be encouraged during professional training.

There are clear benefits to using AI systems to process the content of clinical interviews and digitally translate it into the electronic health record. First and foremost, the doctor has more time to spend with the patient, free from purely administrative tasks, and can even focus his attention and gaze on the person in front of him. However, there is a risk that the systematic use of AI systems in healthcare will depersonalize the clinical relationship, especially if the healthcare system does not provide the time and conditions necessary for doctors and other professionals to provide adequate care to their patients.

How should we respond to this massive deployment of AI in healthcare and society?

First, it is necessary to establish a supranational AI regulatory system based on the risk that AI applications may pose to humans. At the European level, the AI Act aims to ensure respect for fundamental values, especially in healthcare, as this is an area of potentially high risk.⁶ Overall, it is essential that citizens feel confident about AI and its applications in healthcare and research.⁷

On the other hand, comprehensive AI training programs should be implemented in all medical schools and other health faculties and schools. At the professional level, professional associations and organizations should raise their standards in this area.

In short, AI is already being used systematically today, not only in clinical practice, but also in public health and in health management and administration. Only a high level of AI literacy in society in general, and among health professionals in particular, can ensure that the paradigm shift represented by AI contributes to better health care and full human development.