POLAND Trends and Developments Contributed by: Barbara Kiełtyka, Jakub Gładkowski and Małgorzata Kiełtyka, Kieltyka Gladkowski KG Legal
To ensure an AI system is reliable and makes fewer errors, an “adaptive model” is typically introduced, based on the model’s ability to analyse data in real time and make automatic corrections. However, an AI system that updates itself based on new data falls outside the standard clinical evaluation standards of Regulation (EU) 2017/745 (Medical Devices Regula - tion, MDR), as any change may require revalidation. Also, the MDR and Regulation (EU) 2017/746 (In Vitro Diagnostic Medical Devices Regulation, IVDR) do not specify how to classify immersive technologies. Hybrid devices combining AI, robotics and augment - ed reality go beyond the current framework. Therefore, work is under way to amend and adapt the MDR to incorporate rules for hybrid devices, including AI, such that ex-post inspections will be permitted instead of static certification. In 2025, the International Medical Device Regulators Forum published two key documents regulating AI and Software as a Medical Device. These documents contain guidelines on safety, effectiveness and risk assessment, introducing, among other things, con - tinuous market surveillance of the operation of such systems. Furthermore, the EU is working on regula - tions that will define liability standards. AI-Based Drug Discovery and Development Currently, in Polish law, AI-based drug discovery and development rules should be inferred primarily from the provisions of t the Pharmaceutical Law (Polish Act of 6 September 2001 – Pharmaceutical Law (con - solidated text Journal of Laws of 2025, item 750, as amended) and the Regulation of the Minister of Health of 9 November 2015 on the requirements of Good Manufacturing Practice. Before drug design begins, a therapeutic target must be identified – a specific enzyme, a mutated gene or a key signalling path - way. By analysing large biological datasets, including genomic and transcriptomic information derived from next-generation sequencing, AI helps identify the best therapeutic options. This creates regulatory challenges because, according to Polish Pharmaceutical Law, an application for marketing authorisation for a medicinal product must include a description of the medicinal product’s manufacturing process and a description of the control methods used in the manufacturing pro - cess. To this end, the regulatory authority will exam -
ine issues such as analysis of data from multiomics studies, machine learning and molecular dynamics, pharmacokinetic modelling, molecular docking, lead compound identification, and the application of AI in these processes. Therefore, the regulations must be interpreted to account for the production of drugs using AI systems. Healthcare AI Legislative Trends In the field of AI regulation, legislative action is primar - ily undertaken at the EU level to harmonise regulations regarding emerging technologies. Standardising rules for AI systems requires a special approach, as imple - menting AI-based solutions badly can pose significant risks to life, health and safety. Therefore, in healthcare, it is crucial to regulate both technical and algorithmic components that directly impact diagnosis or treat - ment, as well as to ensure a high level of personal data protection, especially sensitive health data. At the EU level, regulations already exist governing medical devices (such as the MDR and IVDR), data flow and protection (Regulation (EU) 2016/679 and Regulation (EU) 2023/2854), and clinical trials (Regu - lation (EU) 536/2014). However, until now, there has been no comprehensive legal act directly addressing the use of AI. The AI Act aims to fill this gap by estab - lishing a single regulatory framework enabling the law - ful and safe implementation of AI systems in the EU single market, while maintaining consistency with the existing EU acquis. Although there is currently no specific regulation dedi - cated solely to AI applications in healthcare, the legal framework established by the AI Act provides the foun - dation for future implementations in this area. Poland, as part of its obligations arising from EU membership, is taking steps to adapt its institutional structures and supervisory mechanisms to the requirements of the AI Act to ensure the effective application of the new regulatory regime in the national healthcare system. Creating Institutions for AI Technologies for Medicine and Supervising AI in Medicine Using Data Digital health authorities already exist in most Member States and address electronic health records, interop - erability, security and standardisation. In carrying out
111 CHAMBERS.COM
Powered by FlippingBook