While artificial intelligence (AI) in medical devices operates within regulatory boundaries, its application throughout a medicine's life cycle presents a less defined landscape.
Amidst the growing adoption of artificial intelligence (AI) and machine learning (ML), the European Medicines Agency (EMA) has released a draft paper delineating its stance on the integration of AI and ML throughout various stages of a medicine's life cycle.
Part of a collaborative effort between the Human Medicines Agency (HMA) and EMA to establish data-driven regulation, the paper underscores the potential benefits AI/ML capabilities offer across all phases of a medicine's life cycle. However, it also emphasizes the need for companies to adhere to legal and ethical standards in their utilization of these technologies.
While the European Union (EU) has developed a comprehensive AI law, the regulatory landscape concerning AI applications in the pharmaceutical industry remains ambiguous. Nonetheless, AI and ML tools hold immense promise across the medicinal product life cycle. From aiding in drug discovery to transforming preclinical development through modelling approaches, these technologies offer innovative solutions. Moreover, AI/ML's data-driven approach enhances clinical trial operations and facilitates product information compilation and pharmacovigilance activities during market authorization and post-authorization stages.
The paper advises companies employing AI/ML technologies to navigate existing legal frameworks carefully, considering potential limitations and challenges such as bias, overfitting, and data protection. Emphasizing a "risk-based approach," it underscores the importance of ongoing engagement with regulators.
Notably, the EMA clarified that it does not regulate AI/ML software used in medical devices. However, it stressed the need for additional requirements when employing CE-marked devices in clinical trials to ensure data integrity, result validity, and subject safety.
Jesper Kjær, Director of the Data Analytics Centre at the Danish Medicines Agency and co-chair of the Big Data Steering Group (BDSG), remarked on the rapidly evolving AI landscape, acknowledging the regulatory challenges it presents. Meanwhile, EMA's Head of Data Analytics and Methods and BDSG co-chair, Peter Arlett, emphasized the importance of collaboration among developers, academics, and regulators to harness the full potential of these innovations for the benefit of patient and animal health.
