Artificial intelligence health: applications, risks, and a complementary future

The report The future of health, which we developed at MAPFRE, points out that digitization in health has gone through two major phases. The first focused on implementing hospital management systems and optimizing administrative processes. Now, a second wave is being driven by AI, which provides significant support to diagnosis and treatment through advanced data analysis and machine learning technologies.

However, the use of this innovation in healthcare suggests major challenges, especially regarding cybersecurity and personal data protection. Given that the health sector handles sensitive information, ensuring the safety of the systems is crucial for their adoption to be successful, as well as training users to use them correctly.

The impact of AI on health is already a tangible reality. Various organizations have begun to implement solutions that optimize medical care and facilitate the patient’s experience, including:

• Virtual assistants and diagnosis with AI: a notable example is the MAPFRE Salud application, which has incorporated an AI-based symptom evaluator that allows the insured to receive a recommendation on the appropriate specialist to visit based on the description of their symptoms. This provides a basis on which the doctor can work on a diagnosis more efficiently and quickly, treating ailments with greater ease.
• Telemedicine and remote rehabilitation: another area where AI is gaining traction is telemedicine. At MAPFRE, in collaboration with Trak, we developed a digital physiotherapy initiative based on the prescription of rehabilitation exercises and AI-assisted patient monitoring. Thanks to this technology, specialists can perform more precise monitoring without needing on-site visits, in turn reducing the rate of abandonment of treatment.
• Process automation: artificial intelligence is improving hospital management by digitalizing documents and optimizing the workflow in medical centers. At MAPFRE, we have automated the capture of information and reimbursement management with generative AI through the MAPFRE Salud application. This makes it possible to improve the customer experience with much better response times and generate operational efficiency.

“AI is a strategic ally that optimizes processes, improves accuracy in diagnoses, and frees up time for professionals to focus on the most important thing: patient care and well-being,” says Pedro Díaz Yuste, CEO of Savia. “Continuing with the example of the MAPFRE Salud application, here AI acts as an intelligent assistant that streamlines medical management, from symptom assessment to automating reimbursements. It makes healthcare more accessible, efficient, and personalized for each patient,” he adds.

The use of artificial intelligence in the health sector represents significant progress, but it also poses cybersecurity risks.

In this regard, healthcare and pharmaceuticals are two of the sectors most exposed to the adverse effects of AI. The digitization of medical records and the automation of processes have made healthcare systems a prime target for cybercriminals—ransomware attacks have been one of the sector’s main concerns, as they can paralyze hospitals and medical centers, compromising patient care. Another risk arising from the application of artificial intelligence in this area is the potential bias in algorithms, the “engine” of AI.

In this regard, the European Union has categorized the health sector as high risk within the AI Act, which implies the need for strict controls on AI-based solutions. The legislation seeks to ensure that algorithms are reliable, transparent, and respectful of patient privacy. Likewise, the regulations indicate that health sector organizations using AI conduct audits that evaluate the accuracy and equity of the models used.

But, despite all the efforts to shield data processing, the risks are not 100% unavoidable: the ECRI 2025 Report on Health Technology Hazards warns of various problems that may arise with the implementation of AI in health. These include biases in the data sets used to train algorithms, which may affect accuracy in diagnosis and treatment. It also highlights “data drift,” a phenomenon in which the quality of AI models degrades over time due to changes in clinical patterns or patient information. To mitigate this risk, continuous learning models are being developed that can be adapted to new data without compromising the reliability of diagnoses.

“Artificial intelligence encompasses more than just crafting functional algorithms; it also entails ethical and human-centered management. At MAPFRE, we adopt ethical AI management that not only complies with regulations, but also places people’s rights at the center and promotes equitable and responsible AI,” comments Diego Bodas, Director of Artificial Intelligence at MAPFRE.

Although AI offers innovative healthcare solutions, its role must be complementary and not a substitute for human judgment. Doctors and specialists continue to play a decisive role in clinical decision-making, and AI must act as a support that optimizes and streamlines processes without replacing professional supervision.

In this context, a hybrid service model can be the key to success in the implementation of AI in health. This approach is based on the combination of technology and the experience of medical professionals.

Although AI systems can improve efficiency and precision in diagnosis, they are not exempt from errors. Validation of results by doctors continues to be a fundamental aspect to avoid critical failures in healthcare. For this reason, explainable AI models are being developed, which allow health professionals to understand how an algorithm has reached a certain conclusion, facilitating a better interpretation of the results.

The future of AI in healthcare is promising, but its adoption must be responsible and reliable. In this regard, for AI to be a real transformation rather than a passing trend, innovation must be accompanied by regulation and governance in data protection and medical oversight, as well as professional training and the development of algorithms under the framework of responsible AI—key to ensuring its ethical and efficient use in the healthcare sector.

RELATED ARTICLES: