AI in the GMP Environment: What do the Regulators Expect?

5 min. reading time | by Azade Pütz
Published in LOGFILE 11/2025

Artificial intelligence is developing faster than any previous technology, promising innovation and competitive advantage. Yet, integrating AI while maintaining safety, quality, and trust poses a challenging task for the pharmaceutical industry. It requires an appropriate regulatory framework as a fundamental condition. Today's editorial outlines which role the EU AI Act, the EMA's reflection paper, and various ISO standards play in shaping the regulatory landscape.

By the way: the GMP Compliance Adviser also offers the use of AI. It features the intelligent search GMP Chat. The GMP Compliance Adviser is the most comprehensive GMP knowledge portal worldwide.

Introduction

The rapid development and increasing use of Artificial Intelligence (AI) in the pharmaceutical industry requires a clear regulatory framework to ensure safety, quality, and trustworthiness. In this complex environment, the EU AI Act, relevant ISO standards, and the EMA's reflection paper on the use of AI in the life cycle of medicines play a central role in shaping the regulatory landscape. These instruments aim to find a balanced approach that promotes innovation while minimising potential risks.

The EU AI Act and its significance for the pharmaceutical industry

The EU AI Act establishes the first comprehensive legal framework for the development, marketing, and use of AI in the EU. This approach is particularly relevant to the pharmaceutical industry, as many AI applications in the sector could be classified as high-risk systems. The act takes a risk-based approach, categorising AI systems from unacceptable to minimal risk.

It sets out strict requirements for high-risk AI systems, which could include many applications in the pharmaceutical industry. These include an appropriate risk management system, high quality of datasets used, comprehensive technical documentation, transparency and provision of information to users, human oversight, and high standards of robustness, accuracy, and cybersecurity. These requirements will have a significant impact on the way AI systems are developed and used in the pharmaceutical industry.

Of note is the demand for transparency and explainability of AI systems. This presents the pharmaceutical industry with the challenge of designing complex AI models in such a way that their decision-making processes are comprehensible and can be interpreted. This is particularly important in medicinal product development and clinical decision support applications, where the traceability of decisions is critical for patient safety and regulatory compliance.

ISO standards as a guide to best practices

Alongside the EU's regulatory efforts, ISO standards play a crucial role in standardising processes and ensuring quality and safety in the pharmaceutical industry.

Although there are no specific ISO standards for AI in the pharmaceutical industry, general AI-related standards such as ISO/IEC 23053:2022 and ISO/IEC 42001 provide important guidelines for the design, implementation, and management of AI systems.

These standards can serve as a bridge between regulatory requirements and practical implementation in industry.

The ISO standards address important aspects such as data management, model development and validation, and risk management. They provide a structured framework for implementing AI systems in accordance with GMP principles. By applying these standards, pharmaceutical companies can ensure that their AI applications are robust, reliable, and compliant with international standards.

The EMAs reflection paper – setting the course in medicinal product development

The EMA reflection paper (Reflection paper on the use of artificial intelligence in the lifecycle of medicines) supplements the regulatory framework with specific considerations on the use of AI throughout the lifecycle of medicines. It highlights the importance of data quality and integrity, transparency, and explainability of AI decisions, and the need for robust validation methods. Of note is the focus on ethical aspects in the development and use of AI, underlining the holistic approach of the document.

An important aspect of the EMA reflection paper is the recommendation for early interaction with regulators, particularly for AI applications with high regulatory impact. This shows that regulatory authorities are taking a proactive and collaborative approach to this rapidly evolving area.

The document also addresses specific areas of application for AI in medicine development, such as its use in clinical trials. It emphasises the need to apply existing guidelines for good clinical practice (GCP) to AI-supported study designs and analyses. This underscores the importance of integrating AI into existing regulatory structures and quality systems.

Special attention is paid to the validation of AI models in GMP environments. Unlike traditional software systems, AI models, especially those based on machine learning, can change behaviour over time. This requires new approaches to continuous validation and monitoring to ensure that the models continue to operate within the validated parameters.

Despite these challenges, integrating AI into GMP processes also offers significant opportunities. AI systems can help to improve process efficiency and control, enable early detection of quality deviations, and optimise resource utilisation. These advantages can lead to a significant increase in product quality and safety while improving the efficiency and competitiveness of companies.

One promising application is the use of AI in process analytical technology (PAT). AI-supported PAT systems can analyse real-time data from the manufacturing process and provide predictions about product quality. This allows proactive process control, helping to reduce the number of nonconforming batches and improve the consistency of product quality.

Summary

The regulatory framework for AI in the GMP environment is developing dynamically. The EU AI Act, relevant ISO standards, and the EMA's reflection paper are specific, complementary cornerstones for the responsible implementation of AI in the pharmaceutical industry. In the future, it will be crucial to harmonise these approaches and adapt them to technological advancements. The industry faces the challenge of balancing innovation and safety.

The development of specific guidelines for the validation and qualification of AI systems in GMP environments will be an important task, taking into account the particular characteristics of AI systems without compromising GMP principles.

Companies that proactively and responsibly integrate AI into their GMP strategies will reap the benefits while maintaining high standards of patient safety and product quality. Success will depend on how well AI technologies are integrated into existing regulatory structures and quality systems. This requires technical expertise and a deep understanding of regulatory and ethical issues. Companies that successfully combine these aspects will be able to realise the full potential of AI in a GMP environment and set new standards in drug development and manufacturing.

Do you have any questions or suggestions? Please contact us at: redaktion@gmp-verlag.de