FDA and EMA Release Joint Guidelines for AI in Drug Development as Regulatory Agencies Advance Framework for Healthcare AI
核心洞察
The FDA and European Medicines Agency (搜索) have jointly established ten principles for artificial intelligence use throughout the medicines lifecycle, from early research to safety monitoring.
The guidelines emphasize human-centric design, risk-based approaches, and comprehensive documentation requirements to maintain drug quality, efficacy and safety standards.
The UK's MHRA has simultaneously launched a call for evidence to inform its National Commission on AI regulation in healthcare, seeking input from healthcare providers and the public through February 2026.
The U.S. Food and Drug Administration (搜索) and the European Medicines Agency (搜索) have jointly released ten foundational principles for artificial intelligence use in drug development, marking a significant step toward international regulatory harmonization as AI technologies increasingly transform pharmaceutical research and development.
The collaborative guidelines provide broad guidance on AI implementation across all phases of medicine development, from early research and clinical trials to manufacturing and post-market safety monitoring. According to the agencies, the principles are designed for medicine developers, marketing authorization applicants, and holders, and will serve as the foundation for future AI guidance in both jurisdictions.
Ten Core Principles for AI in Drug Development
The guidelines establish comprehensive requirements for AI implementation in pharmaceutical development. The principles mandate that AI systems must be "human-centric by design," aligning with ethical values while maintaining appropriate validation, risk mitigation, and oversight through a risk-based approach.
Key technical requirements include adherence to legal, technical, and regulatory standards, with clear definition of AI use contexts and multidisciplinary expert oversight throughout the technology's lifecycle. The guidelines emphasize robust data governance and documentation, requiring comprehensive records of AI's data sourcing, processing, and analytical decisions while maintaining privacy and protection of sensitive information.
The principles also mandate best practices for system design and software engineering that promote transparency, reliability, and patient safety. Risk-based performance assessment must evaluate human-AI interactions and include performance validation through testing, while lifecycle management requires ongoing risk-based quality management, including issue assessment and data drift testing.
Addressing Rapid AI Adoption in Pharmaceuticals
The use of AI technologies across the medicines lifecycle has increased significantly in recent years, according to the agencies. The FDA noted that AI technologies are anticipated to support innovation, reduce time-to-market, strengthen regulatory excellence and pharmacovigilance, and decrease reliance on animal testing by improving prediction of toxicity and efficacy in humans.
As emphasized in the European Commission's Biotech Act proposal, AI holds great promise as a tool to accelerate the path from innovation to safe and effective medicines. The new pharmaceutical legislation accommodates broader AI use in regulatory decision-making and creates additional possibilities for testing innovative AI-driven methods in controlled environments.
UK Regulatory Initiative Parallels International Efforts
Simultaneously, the UK's Medicines and Healthcare products Regulatory Agency has launched a call for evidence on AI regulation in the national healthcare sector. The initiative, running from December 18, 2025, to February 2, 2026, seeks input from healthcare providers, industry, public, and clinicians to inform the newly established National Commission into the Regulation of AI in Healthcare (搜索).
MHRA CEO Lawrence Tallon emphasized the transformative potential of AI, stating: "AI is already revolutionising our lives, both its possibilities and its capabilities are ever-expanding, and as we continue into this new world, we must ensure that its use in healthcare is safe, risk-proportionate and engenders public trust and confidence."
The commission brings together experts from global AI fields, clinicians, patient advocates, and regulators to advise MHRA on the future regulatory landscape for health-related AI. The call addresses clarification of responsibilities among companies, healthcare organizations, individuals, and regulators involved in deploying AI technology, adequacy of current AI healthcare rules, and strategies for maintaining patient safety as AI systems evolve.
Future Regulatory Framework Development
The EMA-FDA initiative builds on collaborative work following the FDA-EU Bilateral meeting in April 2024 and aligns with EMA's mission to promote safe and responsible AI use. The principles will be supplemented by additional EU guidance taking into account applicable legal requirements and relevant new EU legislation in the medicines area.
With ethics at the forefront, the agencies will continue exploring opportunities for global convergence on AI topics to facilitate responsible innovation globally, in close collaboration with international public health partners. The principles-based approach aims to help regulators, pharmaceutical companies, and medicine developers harness AI potential while ensuring patient and animal safety and regulatory compliance.