Clinical AI is Not (Yet) Trustworthy-But It Could Be
By
Ali Saad
Sofia B Dias
Ghada Alhussein
David Lyreskog
Ioannis Gerasimou
Beatriz Alves
Μaarten de Vos
Ioannis Drivas
John Zaras
Andreas Stergioulas
Iskanter Bensenousi
Leontios Hadjileontiadis
Christos Chatzichristos
Stelios Hadjidimitriou
April 29, 2026
Clinical Scorecard: The Trustworthiness of Clinical AI: Current Limitations and Future Potential
At a Glance
Category Detail
Condition Clinical AI Systems Key Mechanisms Trustworthiness, procedural safeguards, ethical alignment Target Population Clinicians, patients, health care institutions Care Setting Health care environments
Key Highlights
Trust in clinical AI encompasses transparency, interpretability, and accountability. The ALTAI framework provides a procedural approach to embedding trustworthiness in AI. Trust is cultivated over time through system behavior and user experience. A persistent implementation gap exists between ethical principles and practical application. AI-PROGNOSIS serves as a case study for applying the ALTAI framework. Guideline-Based Recommendations
Diagnosis
Utilize AI systems with validated trustworthiness metrics. Management
Implement procedural safeguards throughout the AI lifecycle. Monitoring & Follow-up
Continuously assess AI system performance and user trust. Risks
Address ethical and regulatory compliance to mitigate risks. Patient & Prescribing Data
Individuals with conditions requiring predictive modeling, such as Parkinson disease.
AI can generate individualized risk scores and treatment forecasts.
Clinical Best Practices
Embed ethical principles into AI design and governance. Foster transparency and accountability in AI systems. Engage stakeholders in the development and deployment of AI technologies. References
