Hybrid intelligence: understanding how AI reframes risk and uncertainty in dementia care

This article investigates how Artificial Intelligence (AI) systems reshape risk and uncertainty in dementia care by proposing a conceptual shift from technological efficacy to hybrid intelligence. Drawing on empirical research conducted in both residential and home-based care settings in Italy, and based on interviews with professional and informal caregivers, the study explores how AI technologies such as telemonitoring systems and GPS trackers not only assist but co-construct care practices. Rather than functioning as neutral tools, AI systems emerge as epistemic and moral agents that actively participate in framing vulnerability, distributing responsibility, and redefining what counts as actionable risk. The analysis reveals that far from eliminating uncertainty, AI redistributes it, often intensifying the interpretive and affective labour of caregivers. Through the lens of hybrid intelligence, the article foregrounds the relational and situated nature of human-AI collaboration and argues for a critical reconceptualisation of risk in algorithmic health environments. This reframing emphasises the socio-technical entanglements, epistemic asymmetries, and moral decisions embedded in contemporary care infrastructures.