The AI-PROGNOSIS publication “Clinical AI is Not (Yet) Trustworthy-But It Could Be” has been published in the Journal of Medical Internet Research.

The article presents the use of the Assessment List for Trustworthy Artificial Intelligence (ALTAI) within the AI-PROGNOSIS project. AI-PROGNOSIS aims to support Parkinson’s disease diagnosis and care through predictive models and digital biomarkers derived from everyday devices.

The publication maps the seven ALTAI requirements to stages of the AI lifecycle in AI-PROGNOSIS, including design and specification, data preparation, model development and validation, user interface and user experience deployment, external prospective validation, and overall management and workflow.

The article also reports a structured internal survey among AI developers and data scientists involved in AI-PROGNOSIS. Participants rated the relevance of ALTAI subdomains in the project context. Technical accuracy, data governance, and privacy were rated as highly relevant, while societal impact received the lowest prioritization.

The publication presents AI-PROGNOSIS as a context-specific case study on how structured governance frameworks can support the integration of trust-related considerations into clinical AI projects.

Read the full publication:

https://www.jmir.org/2026/1/e85433