GMV tackles algorithmic surveillance and wins the JISA 2026 Prompt Contest

GMV participated in the 33rd National Conference on Innovation and Health in Andalusia with a session on how to prevent harm associated with artificial intelligence in healthcare. In addition, the proposal by Miguel Ferreiro, AI Governance Specialist - Healthcare, won the conference's prompt contest.

Held from October 21–23 in Torremolinos under the theme “Meaningful Relationships,” JISA 2026 brought together industry professionals to discuss innovation and digital transformation in the National Health System.

On October 21, Juan Carlos Sánchez Rosado, Health Business Executive at GMV, and Pol Pérez i Sust, director of the Information Systems Division at the Catalan Health Service (CatSalut), presented Case 1 of Plenary Session 1: Algorithmic Surveillance vs. AI Malpractice. The session examined the potential harm resulting from errors, biases, or unintended behavior of artificial intelligence algorithms in clinical practice.

The risks addressed include prediction errors, overreliance on automated recommendations, loss of clinical judgment, privacy concerns, and inequalities among patient groups. To describe these potential harms, the concept of “AI malpractice” was introduced, and the need to establish continuous monitoring of algorithms—comparable in purpose to pharmacovigilance—was raised.

The presentation featured a case study on supporting referrals from primary care to rheumatology. An unexpected increase in young patients—aged 18 to 35—who were underrepresented in the training data led to incorrect decisions not to refer them. Monitoring made it possible to detect deviations, false negatives, and potential equity issues, and to analyze their causes through collaboration between clinical and technical teams.

The session emphasized that AI safety does not end with its initial validation. It is necessary to validate the models using representative populations, monitor them during actual use, recalibrate them as needed, and establish processes for governance, traceability, and control. The role of standards such as ISO 42001 in integrating AI governance into organizational processes was also highlighted.

GMV's participation in JISA 2026 also earned recognition in the prompt competition, which was designed to solve a problem facing the National Health System (SNS). The proposal by Miguel Ferreiro, AI Governance Specialist - Healthcare, was selected as the winner.

Through its participation in the session and its recognition in the competition, GMV contributed to a key debate on the future of digital health: how to apply artificial intelligence responsibly and ensure it remains safe, equitable, and effective.