The review article published by Núñez Reiz et al.1 makes some valuable remarks on disruptive technologies that will probably revolutionize modern care of critically ill patients. However, we would like to make some considerations on the ethical and legal questions associated with the use of such models for the clinical decision-making process.

The current medical literature agrees that a fundamental component to achieve the safe and effective implementation of these tools of Big Data Analysis (BDA) and Machine Learning (ML) is developing regulatory frameworks to address the unique challenge posed by the actual pace of innovation, the significant risks involved, and the potentially fluid nature of the models of automatic learning2.

In the design of these regulatory frameworks, the remarks made by Núñez Reiz et al. on the privacy and safety of patients whose data are used to build these models are obviously relevant; however, the greater risks involved when implanting these technologies affect, precisely, the critically ill patient who is at the center of the clinical decision-making process with the use of these BDA and ML tools. This means that the patient has a legitimate interest too in the automated management of these data. Actually, the Data Protection General Regulation guarantees the right to obtain sensitive information on the logic used by the algorithm to make this or that prediction on a patient's health status (the making of a profile according to the regulation).

We would like to mention that the individual damage that BDA and ML tools can cause may go misdiagnosed or be irreparable from the very perspective of the individual who is the sole holder of his rights. However, it can also massively affect the fundamental rights of sectors or clusters of society in a relevant way in this collective dimension3. In this sense, the medical literature has shown growing preoccupation due to the reproduction of biases of race or sex in these mechanisms4 that may discriminate these social groups.

Bottom line, the critically ill patient is the paradigm of this problem since exercising one's own rights can prove unfeasible and ineffective in practice. Also, establishing a regulatory framework based on the protection of collective rights with special attention to the processes of validation and supervision of these models or the role of ethics committees the authors talk about can significantly reduce the risks individuals are exposed to when these technologies are implemented.