Regulatory Agencies Establish Principles of Good AI Use in Drug Development
The U.S. Food and Drug Administration's (FDA) Center for Drug Evaluation and Research (CDER) and Center for Biologics Evaluation and Research (CBER) have collaborated with the European Medicines Agency (EMA) to develop a set of 10 guiding principles for good uses of artificial intelligence (AI) in drug development. The principles are aimed at industry members and product developers who are using AI to facilitate and accelerate drug/biological product development.
As the use of AI expands through the drug product life cycle, the two regulatory agencies are working to establish good practice standards for industry and product developers. These standards are designed to align with the rigorous quality expectations of drug development, ensuring patient safety and therapeutic benefit. AI technologies hold significant promise in accelerating time-to-market for new drugs, enhancing pharmacovigilance, and reducing reliance on animal testing—objectives that align with broader FDA initiatives. However, without careful oversight in the design, validation, and application of these technologies, there is a risk of generating inaccurate or unreliable results. By proposing a framework for best practices, the agencies aim to foster the responsible integration of AI in drug development.
The principles also identify areas where international organizations can focus to set further regulatory and legal policies and guidelines for the responsible advancement of AI in drug development and the broader field of medicine.
The 10 principles of good practice for AI use in drug development set out by the FDA and EMA are:
Human-centric by design—for alignment with ethical and human-centric values
Risk-based approach
Adherence to standards—including legal, ethical, technical, scientific, cybersecurity, and regulatory standards
Clear context of use
Multidisciplinary expertise
Data governance and documentation—especially for the privacy and protection of sensitive data
Model design and development practices—promotes transparency, reliability, generalizability, and robustness in the design and development of AI models and technologies
Risk-based performance assessment
Life cycle management—stresses scheduled monitoring and periodic re-evaluation of the systems for performance optimization, including to address potential data drift
Clear, essential information—uses plain language to portray information to the intended audience
To learn more, visit FDA.gov.
ASCO AI in Oncology is published by Conexiant under a license arrangement with the American Society of Clinical Oncology, Inc. (ASCO®). The ideas and opinions expressed in ASCO AI in Oncology do not necessarily reflect those of Conexiant or ASCO. For more information, see Policies.
Performance of a convolutional neural network in determining differentiation levels of cutaneous squamous cell carcinomas was on par with that of experienced dermatologists, according to the results of a recent study published in JAAD International.
“This type of cancer, which is a result of mutations of the most common cell type in the top layer of the skin, is strongly linked to accumulated [ultraviolet] radiation over time. It develops in sun-exposed areas, often on skin already showing signs of sun damage, with rough scaly patches, uneven pigmentation, and decreased elasticity,” stated lead researcher Sam Polesie, MD, PhD, Associate Professor of Dermatology and Venereology at the University of Gothenburg and Practicing Dermatologist at Sahlgrenska University Hospital, both in Gothenburg, Sweden.