The European Medicines Agency (EMA) and the Heads of Medicines Agencies (HMA) have published guiding principles for using large language models (LLMs) across the European medicines regulatory network (EMRN). These guidelines, which form part of the AI workplan to 2028, aim to help staff at regulatory agencies effectively harness the potential of LLMs while addressing challenges like variability, data security risks, and result accuracy. The guiding principles also highlight the importance of protecting personal data while using LLMs at the user and organisational level, and emphasise the need for critical thinking and training. The guiding principles also include general ethical considerations for using LLMs including human dignity, autonomy, rule of law, accountability, security, safety etc, which draw from several sources. A webinar, to be held on 13 September 2024, will introduce these guidelines to the EMRN.