Objective

This call funds projects using AI to improve signal detection and risk prediction in pharmacovigilance (drug safety monitoring), aiming to enhance patient safety and support regulatory decision-making.

Expected Outcomes

Projects must deliver:

• AI algorithms for faster and more accurate detection of adverse drug reactions.

• Predictive models to identify future drug risks early.

• A comprehensive list of data sources and guidance for building a common data model for simultaneous analysis (e.g., EHRs, spontaneous reports, social media).

• Practical recommendations and tools for implementing AI in real-world pharmacovigilance settings.

• Templates, training materials, and education plans to support adoption.

• Strategies for human-in-the-loop (HITL) and human-on-the-loop (HOTL) oversight of AI systems.

• Engagement with regulators (e.g., EMA) to support qualification and uptake of AI methodologies.

Scope

Focus is on applying AI to:

• Detect safety signals from diverse data sources (EHRs, SRSs, registries, social media).

• Predict drug risks before they escalate.

• Generate recommendations for real-world implementation.

• Ensure ethical, legal, and trustworthy use of AI, following frameworks like ALTAI and EU AI regulations.

Exclusions: AI use for case management (ICSRs) and periodic safety reports is out of scope.

Impact

Expected impacts include:

• Earlier and more accurate risk detection

• Better patient safety through timely risk mitigation

• Automated, efficient pharmacovigilance processes

• Lower costs and reduced manual workload

• Regulatory support for AI-driven safety monitoring

• Policy shaping in AI, health data, and emergency response (e.g., EHDS, AI Act, HERA)