
Healthcare professionals, especially doctors, will need to learn how to use artificial intelligence (AI) tools to improve their technical knowledge and maintain necessary updates. And it is no exaggeration to say that in five years, AI will play the role of medical assistant. However, for healthcare experts, the time when technology will replace the emotional intelligence of healthcare professionals and their ability to formulate the correct and most accurate diagnosis is far away.
— When we talk about medicine, there is the pure knowledge and the emotional intelligence of the professional, which is the human relationship. And this relationship does not change — said Paulo Hoff, professor of clinical oncology at the University of São Paulo and president of Oncologia D’Or, during Voices 2025.
Hoff noted that AI tools intended to help doctors, nurses and physical therapists reduce the time spent filling out forms and medical records are in the early stages of implementation. But he insisted that technology will not replace knowledge in validating diagnostics.
There is real concern in the healthcare sector about AI-generated information. For Hoff, it is the doctor’s incommunicable duty to verify the veracity of the responses generated by technological tools.
— In class, we emphasize that AI should be seen as a medical assistant, not a substitute. It extends the capacity of the professional, in terms of documentation and generation of hypotheses, but the legal responsibility remains on the side of the doctor. If the AI makes a wrong diagnosis and the doctor accepts it, it will be his fault – he said.
Used with caution and moderation, AI can help health professionals devote more time to the humanization aspect of care, in addition to increasing the skills of this professional in terms of emotional intelligence, underlined the president of Oncologia D’Or. Paulo Hoff cited tools in which AI is able to understand the natural language of a conversation with patients.
— The tool takes the natural language of the conversation with the patient and configures this content, generating clinical notes and test requests. The professional can pay more attention to the patient and spend less time on technical and bureaucratic aspects. It’s very positive. This is not yet a reality in many practices, but, within the next five years, the majority of them should have access to these types of tools, he said.
AI is also being used in the development of new therapies, particularly in areas such as oncology, cardiology, endocrinology, rheumatology and dermatology. According to Hoff, AI is applied, in a controlled environment, to accelerate the development of new therapies in the digital sphere, without prejudice to all stages of clinical trials under development.
— Currently, out of a hundred pharmaceutical products that enter the development process, three reach the commercialization stage. The others fail at some point during the tests. Hence the enormous financial impact. Using AI tends to increase the success rate and speed up development, he said.
In the field of public health, the use of AI is also expected to soon become a reality, especially in centers already created and equipped with digitalized infrastructure, such as the Cancer Institute of the State of São Paulo and the Hospital das Clínicas of the Faculty of Medicine of the University of São Paulo, Hoff highlighted.
— In the public sphere, in addition to the funding challenge, it is necessary to address issues of interoperability between different medical record systems and laboratories. If these systems do not communicate with each other, the benefits of AI will be very complex. There are also legal questions and the distribution of responsibilities. All this will have to be taken into account,” he said.
Regarding the stage of application of AI in medicine, Hoff noted that not everything will lead to effective results, but other problems may arise unexpectedly.
— We are witnessing the flight of 14 Bis. Of course, not everything we are talking about will happen, but others will happen that we cannot even imagine today — said the doctor.
Another application of AI with a lot of potential, according to Hoff, is linked to the identification of populations more at risk of certain pathologies. The adoption of technology will help reduce the development time of prevention projects. Monitoring and early detection of diseases should also become more efficient processes.