A hybrid deployment model for generative artificial intelligence in hospitals

Maxime Griot, Coralie Hemptinne, Jean Vanderdonckt, Demet YUKSEL

Abstract: This paper analyzes challenges in deploying generative AI in healthcare, examining inadequate evaluation methods, vendor-provided use cases, and adaptation needs. We propose a hybrid framework that uses vendor models for non-medical applications while implementing hospital-managed infrastructure for clinical use cases. This approach balances innovation with patient safety, providing control over adaptation while mitigating biases, ensuring regulatory compliance, and maintaining long-term stability for sustainable healthcare integration.