Purpose: This contribution explores the underuse of artificial intelligence (AI) in the health sector, what this means for practice, and how much the underuse can cost. Attention is drawn to the relevance of an issue that the European Parliament has outlined as a "major threat" in 2020. At its heart is the risk that research and development on trusted AI systems for medicine and digital health will pile up in lab centers without generating further practical relevance. Our analysis highlights why researchers, practitioners and especially policymakers, should pay attention to this phenomenon. Methods: The paper examines the ways in which governments and public agencies are addressing the underuse of AI. As governments and international organizations often acknowledge the limitations of their own initiatives, the contribution explores the causes of the current issues and suggests ways to improve initiatives for digital health. Results: Recommendations address the development of standards, models of regulatory governance, assessment of the opportunity costs of underuse of technology, and the urgency of the problem. Conclusions: The exponential pace of AI advances and innovations makes the risks of underuse of AI increasingly threatening. Graphical

Artificial intelligence (AI) has been identified as a technological revolution for the next decade (Kelly et al., 2023), with the potential to improve human life in several domains. The present study analyses the factors influencing AI acceptance in adults. A quantitative synthesis based on Kelly et al. (2023) was conducted using one-stage meta-analytic structural equation modeling (OSMASEM; Jak et al., 2021).A total of 40 studies comprising 36,403 subjects were included, yielding 43 correlation matrices and 110 correlation coefficients. The OSMASEM results support a dual-pathway model to AI acceptance: perceived usefulness (PU) is a strong predictor of trust in AI (TR) (β31 = .75, p < .001), which in turn predicts behavioral intention (BI) (β53 = .43, p < .001). Concurrently, performance expectancy (PE) predicts attitudes towards AI (ATT) (β42 = .75, p < .001), which also predicts BI (β54 = .40, p < .001). TR and ATT mediate positive relationships between PU (β51 = .32, p < .001) and BI, and between PE (β52 = .30, p < .001) and BI.This review contributes to understanding the factors influencing AI acceptance. In practical terms, these findings emphasize the importance of internal factors such as TR and ATT, alongside conventional acceptance factors (PU, PE). The need for further research in this area is highlighted, with the recommendation to employ specific AI acceptance models.These results suggest that while core cognitive evaluations (PU, PE) remain crucial, their influence is channeled through more proximal affective and relational constructs (TR, ATT) indicating a potential evolution from traditional technology acceptance frameworks when applied to AI.