Principles of Good AI Practice in Drug Development: EMA and FDA Set a Common Framework

At the end of 2025, the DrugCard team hosted the first-ever practical pharmacovigilance conference in Kyiv, Ukraine, focused on applying artificial intelligence (AI) in drug safety. This landmark event highlighted that DrugCard is at the forefront of pharmacovigilance innovation, integrating AI into monitoring and safety processes. The recent announcement that EMA and FDA have established principles of good AI practice in drug development further confirms the growing role of AI in medicine. These principles guide pharmaceutical developers, regulators, and other stakeholders to use AI safely, ethically, and effectively across the entire drug lifecycle.

Why Principles of Good AI Practice in Drug Development Matter

The use of AI in drug development has increased rapidly in recent years. As emphasised in the European Commission’s Biotech Act proposal, AI has the potential to accelerate the development of safe and effective medicines.

The principles of good AI practice in drug development help manage risks, ensure reliable and transparent data, and enhance predictive capabilities in clinical trials, manufacturing, and post-marketing safety monitoring. They also support reducing animal testing by improving the prediction of drug safety and efficacy in humans.

By providing a common framework, these principles promote innovation while maintaining high standards of patient safety.

EMA and FDA’s 10 Principles of Good AI Practice

EMA and FDA outlined ten guiding principles to ensure AI is applied responsibly across the drug lifecycle:

Human-centric by design: AI aligns with ethical and human-focused values.

Risk-based approach: Validation, oversight, and mitigation depend on context and model risk.

Adherence to standards: Compliance with legal, ethical, technical, scientific, and regulatory standards (GxP).

Clear context of use: Well-defined role and scope for AI technologies.

Multidisciplinary expertise: Integration of AI and domain knowledge throughout the lifecycle.

Data governance and documentation: Transparent, traceable, and secure data handling.

Model design and development practices: Reliability, explainability, and predictive performance.

Risk-based performance assessment: Evaluation of AI-human interactions using context-appropriate metrics.

Life cycle management: Continuous monitoring, re-evaluation, and addressing potential issues.

Clear, essential information: Accessible explanations of AI functionality, limitations, and updates for users and patients.

These principles establish a framework for good practice in AI use, encourage responsible innovation, and strengthen pharmacovigilance worldwide.

Why Principles of Good AI Practice in Drug Development Matter for Pharmacovigilance

While the principles of good AI practice in drug development apply throughout the lifecycle of medicines, they are particularly relevant to pharmacovigilance. Drug safety monitoring relies on accurate, timely, and reliable data, and AI can significantly improve the detection, assessment, and management of safety information.

By following these principles, pharmacovigilance teams can use AI as an intelligent assistant to make their work faster and more accurate:

• Process data faster – AI can quickly analyse large amounts of safety information.

• Support decision-making – AI highlights important patterns and helps experts focus on critical cases.

• Reduce errors – Automated analysis minimises mistakes and improves reliability.

• Assist experts – AI acts as a helper, allowing human teams to concentrate on complex judgments.

In short, applying these principles in pharmacovigilance turns AI into a reliable assistant, speeding up data handling and supporting safer, more intelligent decisions.

How DrugCard Leads Innovation Following These Principles

At DrugCard, we actively apply AI in pharmacovigilance. The principles of good AI practice in drug development confirm that our approach is both responsible and innovative.

By integrating these principles, DrugCard:

• Reduces risk in drug safety monitoring

• Increases efficiency in pharmacovigilance workflows

• Accelerates medicine development while maintaining patient safety

Aligning our work with EMA and FDA guidance ensures that AI is applied ethically, reliably, and effectively, reinforcing DrugCard’s leadership in AI-driven pharmacovigilance solutions.

We continue to implement advanced AI-driven solutions that reduce risks, increase efficiency, and accelerate pharmacovigilance processes, while maintaining the highest standards of patient safety and regulatory compliance.