Work Packages
The project contains 5 Work Packages, which you can read more about here.
Lead Beneficiary
NTNU - Norwegian University of Science and Technology
Objectives
To ensure efficient and effective implementation of the project, including coordination between partners, risk management, and communication with the funding authority.
Description
This WP covers the overall management of RISK-AI. It includes planning and monitoring of activities, financial and administrative follow-up, and coordination of the advisory board with experts in ethics, regulation and medicine.
WP1 also oversees project risk management and ensures timely delivery of milestones and reports to NordForsk and other stakeholders.
Lead Beneficiary
NSE - Norwegian Centre for E-health Research
Partners Involved
NTNU - Norwegian University of Science and Technology, UniOulu - University of Oulu
Objectives
To map and analyse the regulatory and governance landscape for AI in healthcare in the Nordic–Baltic region, with a focus on the EU AI Act, and to identify key stakeholders and governance models for AI risk management.
Description
This WP examines how the AI Act and related policies shape organisational responses to AI in healthcare. It includes systematic mapping of policy documents, legal requirements and governance structures, and interviews with policymakers and other stakeholders across the region. The findings provide a regulatory and governance baseline that feeds directly into the conceptualisation of Responsible Risk in WP3 and the capacity-building activities in WP4.
Lead Beneficiary
NTNU - Norwegian University of Science and Technology
Partners Involved
All partners
Objectives
To conceptualise Responsible Risk, define thresholds for acceptable clinical risk and coping strategies in AI-supported healthcare, and integrate this concept into the MAS-AI framework for assessment of AI technologies RISK-AI Submission.pdf.
Description
WP3 develops the core theoretical and practical foundation of RISK-AI. Building on literature, policy analysis and stakeholder interviews, the WP conceptualises Responsible Risk at macro (policy), meso (organisational) and micro (clinical/patient) levels.
It then adapts the MAS-AI framework to incorporate Responsible Risk into its ethical and legal dimensions and pilot tests the updated framework in two clinical use cases in Norway and Latvia, assessing AI technologies at different stages of their implementation lifecycle.
A dedicated PhD position is embedded in this WP.
Lead Beneficiary
UniOulu - University of Oulu
Partners Involved
OUH - Odense University Hospital
Objectives
To build capacity among policymakers, healthcare professionals and other stakeholders to understand, assess and manage AI-related risks, and to develop practical tools and training materials based on Responsible Risk and MAS-AI.
Description
WP4 translates insights from WP2 and WP3 into concrete educational and practical resources. Through co-design with stakeholders, it identifies competence needs and develops training formats, guidance material and a toolkit (e.g. checklists, templates, decision support) for AI risk assessment and governance. These resources are piloted and refined in the Nordic–Baltic context to support trustworthy adoption of AI in healthcare.
Lead Beneficiary
OUH - Odense University Hospital
Partners Involved
All partners
Objectives
To ensure effective dissemination of project results and active engagement of key stakeholders through publications, digital channels, policy briefs, podcasts and co-creation formats, and to support uptake of project findings in practice, policy and education.
Description
This WP designs and implements the project’s communication and dissemination strategy.
Activities include stakeholder mapping and engagement, creation of the project website and visual identity, production of podcasts on AI governance and Responsible Risk, and development of policy briefs and toolkits.
WP5 also explores the use of a domain-trained large language model to provide AI-assisted access to outputs from WP2–4, enabling broader and more user-friendly access to the project’s knowledge base.