Last November, the European Medicines Agency (EMA) and the Heads of Medicines Agencies (HMA) convened a multi-stakeholder workshop to discuss the evolving role of Artificial Intelligence (AI) in medicines regulation. The discussions confirmed that AI integration across the EU regulatory network is accelerating — with significant strategic implications for the pharmaceutical industry.
Below are the key takeaways.
- Regulatory Transition to Operational AI
EU regulators — including European Commission (EC) — are moving from AI exploration to operational deployment within regulatory workflows.
This transition is framed by a lifecycle-based, risk-driven approach under:
- The Artificial Intelligence Act
- The European Health Data Space (EHDS)
- Existing sectoral pharmaceutical legislation
Clear guardrails are being established to ensure responsible and compliant AI use.
- AI Already in Use by Authorities
National competent authorities are already applying AI in several high-impact areas:
- Pharmacovigilance signal detection using real-world data
- Clinical data analysis
- Regulatory document consistency and quality control
- Chemical similarity assessments and predictive modelling
Across all applications, human oversight and validation remain central principles.
- GxP Validation and Compliance Expectations
Industry representatives emphasized the need for structured validation frameworks for AI systems operating in regulated environments.
The proposed “onion model” approach includes validation at three levels:
- System
- AI subsystem
- Model
Key compliance components include:
- Clearly defined intended use
- Risk-based validation
- Continuous performance monitoring
- Bias detection and mitigation
AI governance is rapidly becoming a core GxP compliance priority.
- AI Act and Emerging Regulatory Guidance
The EU AI Act introduces mandatory requirements for high-risk AI systems, aligned with existing medicinal product legislation.
Important regulatory developments highlighted during the workshop include:
- GMP Annex 22 (AI in critical GMP applications)
- International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) M15 (draft)
- EMA–FDA Guiding Principles for Good AI Practice
- AI terminology and glossary harmonisation initiatives
Regulatory expectations are becoming more structured — and more concrete.
- Ethical and Transparency Imperatives
The use of AI in clinical trials and patient-facing applications raises important concerns regarding:
- Fairness
- Explainability
- Transparency and disclosure
Regulators and stakeholders stressed the importance of documented AI use and robust ethical governance frameworks to maintain trust and accountability.
- Strategic Priorities Identified
Three strategic priorities emerged from the discussions:
- International harmonisation
- Accelerated and clarified regulatory guidance
- Regulatory sandboxes to enable responsible experimentation
Conclusion
AI integration in medicines regulation is no longer theoretical — it is operational and expanding.
For pharmaceutical companies, early regulatory engagement, robust validation frameworks, and proactive alignment with the AI Act are becoming strategic imperatives. Organizations that embed structured AI governance now will be better positioned to ensure compliance, innovation, and long-term regulatory readiness in an increasingly AI-enabled environment.
At Asphalion, digitalisation is at the core of our processes, embedded across our regulatory and quality activities, enabling greater efficiency, strengthened oversight, and sustainable compliance in a rapidly evolving landscape.
Contact us for further information: [email protected]
