FDA-EMA Good AI Practice Principles: What They Mean for Your Medical Team

On January 14, 2026, the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) jointly published the "Guiding Principles of Good AI Practice in Drug Development," ten high-level principles aimed at guiding the safe and responsible use of artificial intelligence across the entire drug lifecycle ^[1]. It is the first joint document of its kind between the two agencies, and while it doesn't constitute formal regulatory guidance, it signals where AI regulation in pharmaceutical development is heading on both sides of the Atlantic.

The FDA's original document defines the scope precisely: the principles apply to system-level AI technologies used to generate or analyze evidence across the pharmaceutical product lifecycle, including nonclinical, clinical, post-marketing, and manufacturing phases ^[1]. The conceptual starting point is clear: medicines are authorized based on demonstrated quality, efficacy, and safety, and the use of new technologies such as AI must reinforce those requirements, not compromise them ^[1].

A legal analysis of the joint publication summarizes the core concepts around which the ten principles are organized: alignment with ethical, human-centered values; adherence to current drug development standards; a risk-based approach; a clearly defined context of use; robust data governance; multidisciplinary expertise; lifecycle management of the AI system; and clear, accessible communication about how it works ^[2].

For a medical or regulatory affairs team already using — or evaluating — AI for tasks like evidence mapping, educational content generation, or medical data analysis, these principles offer a useful common language even before they become a formal regulatory requirement: document the context of use for each AI tool, maintain multidisciplinary human oversight, and be able to explain, in clear terms, what the system does and with what limitations. Adopting that standard early narrows the compliance gap once formal guidance arrives.