EU agencies adopt workplan on AI in medicines regulation

The European Medicines Agency (EMA) and the Heads of Medicines Agencies (HMA) have adopted a workplan to facilitate the “responsible and beneficial” use of artificial intelligence (AI) in medicines regulation, while managing the technology’s risks. The workplan was developed by the agencies’ Big Data Steering Group and is set to govern AI activities through 2028.

“Artificial Intelligence (AI) systems are gradually becoming a dominant tool supporting intellectual work and powering automation across all walks of life,” states the workplan. “The European medicines regulatory network’s (EMRN) vision on AI is for 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.” The workplan approved by EMA’s Management Board at its December meeting.

According to an announcement, while pharmaceutical companies are increasingly using AI tools in their research, development and monitoring of medicines, national competent authorities (NCAs) are also starting to use and develop these tools. 

The workplan addresses four “critical dimensions” to facilitate the development and use of “responsible and beneficial AI.: These areas include guidance, policy and product support; AI tools and technology; collaboration and training; and experimentation.

In the guidance and policy area, the workplan stated that “continued support to the development and evaluation of AI in the medicines’ lifecycle will be provided.” For example, comments received on the EMA’s AI reflection paper will guide the development and evaluation of AI uses. For example, comments received in the pharmacovigilance area will address whether this topic will be addressed in a single guidance or multiple guidances. The deadline for comment closes on 31 December. (RELATED: EMA details lifecycle approach to AI/ML drug development in new reflection paper, Regulatory Focus 20 July 2023)

From mid-2024 onwards, regulators will be preparing for the AI Act coming into force. On 8 December, the European Parliament and European Council reached a provisional agreement on a plan to govern the regulation of AI in the EU.

In the tools area, regulators said that initial knowledge mining tools will be rolled out for the EMRN in the first quarter of 2024, while a phased roll-out and monitoring of large language models (LLMS) and related chatbots will be released in the second quarter of 2024.

The use of LLMs and chatbots, “will become a dominant tool in intellectual work as personal assistants.”

Regulators also expect to complete a survey on the LMRN network’s capability to analyze data using AI in the third quarter of 2024.

The LMRN network also plans to work with regulatory partners on AI collaborative activities; such partners include the International Coalition of Medicines Regulatory Authorities (ICMRA), other EU agencies, the devices sector and academia on AI.

In the experimentation area, there are plans to develop internal guiding principles for responsible AI; this activity will be supported by the European Specialised Expert Community (ESEC). As part of this work, regulators plan to issue technical deep dives in such areas as digital twins. These are expected to be completed in the second quarter of 2024 and the second quarter of 2026.

EU AI announcement, Workplan