Artificial Intelligence in European Medicines Regulation: From Vision to Action. Harnessing the Capabilities of Artificial Intelligence for the Benefit of Public and Animal Health

The paper “Artificial intelligence in European medicines regulation” (Nature Reviews Drug Discovery, 2022),¹ presented the European Medicines Agency's (EMA) perspective on artificial intelligence (AI). The paper detailed an approach to working together with stakeholders to deliver a roadmap on AI for the benefit of public and animal health.

Through the joint Heads of Medicines Agency (HMA) and EMA Big Data Steering Group (BDSG),² the European Medicines Regulatory Network (EMRN) has since made rapid progress.

Here, we set out the vision and areas of focus, and how they translate into a multi-annual workplan aimed at enabling the safe and responsible use of AI in the medicines lifecycle for the benefit of public and animal health.

There are three areas of focus for the EMRN's AI transformation: the ability to regulate products that include AI in their lifecycle as required, the ability to leverage AI for process improvement and analytics, and the ability to leverage AI for advanced healthcare data analytics. These focus areas underpin the EMRN's vision for AI: “a regulatory system harnessing the capabilities of AI for personal productivity, process automation and systems efficiency, increased insights into data and strengthened decision-support for the benefit of public and animal health.”³

In November 2023, the EMRN held a public workshop⁴ to hear the views of stakeholders on a draft reflection paper on the use of AI in the medicines lifecycle,⁵ on a vision for AI in medicines regulation and on a draft plan of actions to deliver that vision. With the perspectives of stakeholders shared, in December 2023 the first EMRN multi-annual AI workplan was published.³

The workplan includes four interconnected streams. “Guidance, policy, and product support” will ensure there is support to product development and submissions through advice to biopharmaceutical companies and through guidance on AI, and that the EMRN adapts quickly to the evolving AI legal framework. “Tools and technology” will ensure robust technology is available to the EMRN to enable the deployment of AI-powered applications in full compliance with EU data protection requirements. “Collaboration and change management” will ensure the input of stakeholders at European and international levels is leveraged, and EMRN staff is empowered with the knowledge and skills needed to realize the benefits and manage the risks of AI. Through fostering a culture of continuous learning and adaptation, the EU Network Training Centre (EU NTC) endeavors to cultivate a dynamic learning ecosystem that thrives on innovation, inclusivity, and sustainability in a rapidly evolving landscape of AI. With respect to “Experimentation,” the EMRN aims to create an environment that can explore AI's potential while mitigating risks related to privacy, bias, and accountability, as well as to understand the maturity of the technology and to avoid the pitfalls of the technological imperative, such as rushed implementations of AI. This approach underscores a commitment to harnessing AI's benefits while safeguarding against potential harms.

Each stream supports the others through feedback loops. For instance, under “Experimentation” the EMRN aims to create a research agenda that will also guide the EMRN's approach to “Tools and technology” and inform on “Guidance and policy development.” Also, the “Collaboration and change management” stream seeks to empower regulators to effectively navigate technical and scientific aspects of machine learning through training programs and collaborative platforms, which in turn will support the evaluation of AI in the medicines lifecycle.

The four streams of the workplan will deliver on all three areas of focus. Some deliverables aim at regulating the application of AI systems in relation to the lifecycle of medicinal products, while others are internally focused on leveraging the use of AI within medicines regulation. Collectively, the four streams will contribute to animal and public health by ensuring a clear regulatory pathway for drug development, by increasing the efficiency of regulatory processes, and by further improving the quality of decision making on the benefits and risks of drugs. These streams provide a framework that allows the European Medicines Regulatory Network to have a strategic approach to AI that embeds key ethical and patient-centric values⁶ as well as cooperation with stakeholders. Simultaneously, the EMRN will continue to consider the need to increase capability and capacity to realize the vision and how to best deliver these goals.

Highly capable general-purpose large language models (LLM) have become widely available and are supporting an increasingly large number of applications. The global introduction and rapid widespread use of LLMs illustrate the fast pace of change of science and technology and the EMRN AI approach needs to account for this.

The EMRN AI workplan, including timelines, is available online at https://www.ema.europa.eu/en/documents/work-programme/multi-annual-artificial-intelligence-workplan-2023-2028-hma-ema-joint-big-data-steering-group_en.pdf and will be reviewed and updated at least annually, overseen by the BDSG. The BDSG will establish an observatory to inform its work, will learn from the experimentation across the network, and will maintain open dialogue with stakeholders in the EU and internationally.

Stakeholder engagement and collaborations are a core part of the workplan as they expedite learnings and promote certainty and predictability in a fast-changing environment. Since the publication of the EMRN multi-annual workplan on AI, FDA has published its approach “Artificial Intelligence & Medical Products: How CBER, CDER, CDRH and OCP are working together.”⁷ The publication reveals significant similarities with the EMRN focus areas and workplan, particularly in terms of supporting drug development, stakeholder outreach and collaboration, and experimentation. This illustrates that there is much to be gained from regulatory collaboration including on guidance development, on priorities for experimentation and on sharing of experience. The EMRN is committed to fostering such international collaboration.

No funding was received for this work.

The authors declared no competing interests for this work.

The views expressed in this article are the personal views of the author(s) and may not be understood or quoted as being made on behalf of or reflecting the position of the regulatory agency/agencies or organizations with which the author(s) is/are employed/affiliated.