What is it about?
Artificial intelligence (AI) is being used more and more in healthcare to improve diagnosis, treatment, and personalised care. However, if not designed carefully, these technologies can unintentionally increase existing inequalities and exclude certain groups from their benefits. In our study, we brought together experts from healthcare, ethics, law, social sciences, and patient advocacy to explore how AI in healthcare can be made more inclusive. During a five-day co-creation workshop in the Netherlands, we worked together to identify key issues and come up with practical solutions. We focused on three main areas: ensuring diverse voices are heard during the development of AI tools, making data collection fair and representative, and creating regulations that protect human rights. From these discussions, we created the PREFER-IT framework, which outlines eight key principles for inclusive health AI: Participatory and co-design approaches; Representative and diverse data; Education and digital literacy; Fairness; Ethical and legal accountability; Real-world validation and feedback; Inclusive communication; and Technical interoperability. This framework helps guide developers, policymakers, and healthcare professionals in creating AI systems that are not only effective but also fair and respectful of all users. Our work emphasises the importance of involving patients and communities in shaping the future of AI.
Featured Image
Photo by Zacqueline Baldwin on Unsplash
Why is it important?
Most guidance on fair AI in healthcare either focuses on technical fixes, such as correcting biased algorithms, or stays at the level of broad ethical principles. PREFER-IT connects the two by treating inclusion as a design principle from the start of a project rather than a check at the end. It places each principle on two axes. The first is the parts of an AI system: people, data, processes, the system itself, and governance. The second is each AI lifecycle phase, from design to monitoring. Patients helped shape both the problems and the solutions during the workshop, so the framework draws on lived experience as well as expert knowledge. Developers, clinicians, hospitals, funders and policymakers can use it as a planning tool. They can also build it into checkpoints that already exist, such as ethics review, data protection impact assessments and funding criteria. The timing matters because the EU AI Act is bringing in new requirements on AI literacy and accountability.
Perspectives
What stayed with me most from this work was the second day of the workshop, when patient experts joined the groups. Once they shared what it is like to navigate the healthcare system, the groups rewrote problem statements they had agreed on the day before. For me, that was clear evidence that inclusion changes outcomes and not only processes. Bringing together such a large and varied group of co-authors was a challenge in itself, and it mirrored the message of the framework: bringing different voices together takes time and effort, but the result is richer. I hope teams actually pick PREFER-IT up and adapt it, and I look forward to testing it in real projects and in settings beyond Europe.
Patrícia Pita Ferreira
Rijksuniversiteit Groningen
Read the Original
This page is a summary of: PREFER-IT: A transdisciplinary co-created framework to realise inclusive health AI, PLOS Digital Health, September 2026, PLOS,
DOI: 10.1371/journal.pdig.0001096.
You can read the full text:
Contributors
The following have contributed to this page
