05/26/2026
Photo By Aaron Burden on Unsplash
On January 14th, 2026, the US Food and Drug Administration’s (FDA) Center for Drug Evaluation and Research (CDER) and Center for Biologics Evaluation and Research (CBER) collaborated with the European Medicines Agency (EMA) to develop 10 Guiding Principles of Good AI Practice in Drug Development. When followed, these principles help to ensure data reliability, patient safety, and regulatory excellence while using AI in trial conduct. They are intended to be a foundational basis for developing good practices with AI, addressing the unique nature of AI, as well as cultivating its future growth.
The 10 principles are as follows:
1. Human-centric by design – This means that the development and use of AI should align with ethical and human-centric values.
2. Risk-Based Approach – AI development and use should follow a risk-based approach with proportionate validation, risk mitigation, and oversight. This is similar to the risk-based approach that is encouraged in the ICH E6(R3).
3. Adherence to Standards – This includes relevant legal, ethical, technical, scientific, cybersecurity, and regulatory standards, including Good Practices (GxP).
4. Clear Context of Use – A well-defined context of use includes the role and scope for why it is being used.
5. Multidisciplinary Expertise – Expertise from various fields within the clinical research ecosystem should be integrated throughout the technology’s life cycle.
6. Data Governance and Documentation – The governance of data should include privacy and protection of sensitive data throughout its lifecycle while data documentation should be detailed, traceable, and verifiable for all data source provenance, processing steps, and analytical decisions. This principle reflects principles found in the EU Commission’s recently updated Annex 11: Computerized systems.
7. Model Design and Development Practices – Development of AI technologies should follow best practices in model and system design and software engineering as well as leverage data that is fit-for-use and considers:
Interpretability,
Explainability,
And Predictive Performance.
Good model and system development for AI contributing to patient safety should also promote:
Transparency,
Reliability,
Generalizability,
And Robustness.
8. Risk-Based-Performance Assessment - Assessments should evaluate the complete system including human-AI interactions, using fit-for-use data and metrics appropriate for the intended context of use, supported by validation of predictive performance through appropriately designed testing and evaluation methods.
9. Life Cycle Management - Risk-based quality management systems are implemented throughout the AI technologies’ life cycles, including to support capturing, assessing, and addressing issues. The AI technologies undergo scheduled monitoring and periodic re-evaluation to ensure adequate performance (e.g., to address data drift).
10. Clear, Essential Information - Plain language is used to present clear, accessible, and contextually relevant information to the intended audience, including users and patients, regarding the AI technology’s context of use, performance, limitations, underlying data, updates, and interpretability or explainability.
This collaboration between the FDA and EMA on these principles will empower stakeholders to advance responsible innovations in this area. The full list can be found on the FDA’s website. The FDA also released a Q&A on “Artificial Intelligence for Drug Development” that includes additional information about the FDA’s thoughts on AI in drug development as well as a list of related guidance and resources. If you have not already, please see the previous 2 parts of this series (Part 1, Part 2) for more information on the FDA’s use of AI and sign up for our free blog and newsletter to stay up to date on clinical trial related news such as this.
-The Clinical Pathways Team
Enjoy this blog? Please like, comment, and share with your contacts.