EMA and FDA agree 'principles' for AI in drug development
The EMA and FDA have published a set of 10 'common principles' that should be taken into account when using AI in the development of medicines.
The list has emerged as the adoption of AI among biopharma companies is rapidly accelerating, upending usual practices across the product life cycle from drug discovery through to clinical testing and commercial rollout.
According to the two regulators, the list (PDF) will underpin future guidance on AI in the US and EU and support international standard-setting. It gives "broad guidance on AI use in evidence generation and monitoring across all phases of a medicine, from early research and clinical trials to manufacturing and safety monitoring."
Top of the list is that AIs must be human-centric by design and align with ethical values. They must be developed using a risk-based approach that takes into account the context of their use, and adhere to current legal, ethical, technical, scientific, cybersecurity, and regulatory standards.
They must have a clear context of use and be developed with multidisciplinary expertise, under strict data governance and documentation controls, including on privacy and protection of sensitive data, and using best practices in model and system design and software engineering.
Risk-based performance assessments, including human-AI interaction testing, will be required before they can be used, along with ongoing monitoring and periodic re-evaluation to ensure they remain fit for purpose. Finally, the regulators want the AIs to use plain language that is "clear, accessible, and contextually relevant information to the intended audience, including users and patients."
To realise the benefits of AI will require the technology to be managed, including the mitigation of risks, said the regulators.
"The guiding principles of good AI practice in drug development are a first step of a renewed EU-US cooperation in the field of novel medical technologies," commented Olivér Várhelyi, European Commissioner for Health and Animal Welfare.
"The principles are a good showcase of how we can work together on the two sides of the Atlantic to preserve our leading role in the global innovation race, while ensuring the highest level of patient safety," he added.
The rapid adoption of AI in the industry can be seen from the high frequency of deal-making between biopharma groups and AI specialists.
The last few days alone have seen partnerships announced involving Iktos and both Servier and Pierre Fabre, Insilico Medicines with Servier, and GSK with Noetik and Helix, while AstraZeneca acquired Boston-based partner Modella AI to bring its foundation models and AI agents in-house and support its R&D operations.
Photo by Zulfugar Karimov on Unsplash
