Molecules & Machines: The Rise of AI-Assisted Drug Development

Artificial intelligence (AI) is transforming industries across the globe, and the pharmaceutical sector is no exception. By harnessing AI’s ability to process and analyze massive datasets, pharmaceutical companies are unlocking new opportunities to accelerate drug discovery, optimize clinical trials, and improve manufacturing efficiency. Yet, as AI’s influence in drug development grows, so do the legal, ethical, and regulatory questions it raises. This article explores how AI is reshaping drug development, the challenges and opportunities it presents, and the emerging regulatory landscape that will define its use.

THE ROLE OF AI IN DRUG DEVELOPMENT

AI tools can significantly accelerate drug development by modeling how compounds interact with biological systems and identifying potential candidates for further research. These tools process vast datasets to predict which compounds will achieve desired effects on cellular processes and human health. Specifically, AI-based tools can be used to identify optimal target protein structures and help predict potential efficacy and safety characteristics of drug candidates through advanced modeling of molecular interactions and biological pathways. In addition to shortening early development timelines, this novel approach could also enable more targeted therapeutic strategies.

Data Privacy and Intellectual Property Considerations

Using de-identified patient data for model training without appropriate consent raises legal and ethical concerns. Proprietary AI models often lack transparency, complicating regulatory reviews to the extent that AI-generated data is submitted to regulators. As such, intellectual property (IP) agreements and data-sharing protocols must be carefully structured to ensure ethical use and alignment with privacy, regulatory, and compliance obligations.

There are also important IP considerations associated with AI-assisted discovery. For example, if the AI model is trained on a pharmaceutical company’s proprietary dataset, the company may want to (1) receive payments, IP rights, or other economic benefits for the use of its data, and/or (2) restrict the AI model provider from using those learnings or improvements for the benefit of a competitor. This is particularly true if the company’s dataset is uniquely valuable and/or confidential. AI vendors should, in turn, anticipate questions about whether similar outputs could be generated for third parties using distinct inputs.

In order to preserve freedom to operate and work with other customers, AI model providers should determine whether, and be ready to explain how, they might generate essentially the same output for a third party based on such a third party’s inputs (without using this particular pharmaceutical company’s results or other confidential information).

FDA Oversight and Regulatory Framework

The US Food and Drug Administration (FDA) is taking a closer look at how AI is used in drug development and is working to adapt its regulatory framework to keep pace with these innovations. While AI offers exciting possibilities for speeding up discovery and improving drug safety, FDA is clear that these tools must meet the same high standards as traditional methods.

In recent years, FDA has published several discussion papers and draft guidance documents to outline its thinking. In May 2023 (and recently updated in February 2026), FDA released a discussion paper and request for feedback from the industry on Using Artificial Intelligence & Machine Learning in the Development of Drug & Biological Products. This paper outlined FDA’s view that AI and machine learning could accelerate drug discovery, manufacturing, clinical trials, safety surveillance, and overall product quality. It emphasized key principles like reducing bias in AI models, ensuring data quality, and making AI systems understandable enough for regulators to assess their impact. Although this document is not formal guidance from the agency, it represents FDA’s early thinking and is therefore likely to inform future policy.

In January 2025, FDA released draft guidance titled Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products. This guidance outlines a structured context-of-use framework, encouraging sponsors to define how AI models will be used (e.g., for predicting safety signals, optimizing manufacturing parameters, or supporting regulatory submissions) and tailor assessments accordingly. The guidance emphasizes that sponsors should evaluate model validity, verification, and applicability based on the model’s risk level and maintain robust documentation to demonstrate that the AI tool can reliably support regulatory decisions.

Recognizing the adaptive nature of AI models, FDA stresses the importance of transparency, post-approval updates, and ongoing monitoring to detect performance drift or emerging bias. Importantly, this draft guidance applies when AI outputs inform drug safety, efficacy, or quality; it does not extend to purely operational or back-office uses, nor does it cover drug discovery tools used internally.

FDA is reviewing public input on its draft guidance and may issue final guidance in the relatively near future. The agency is also considering broader applications of AI in post-market surveillance, labeling, and toxicity prediction, aligning with its goal to reduce reliance on animal testing, and is also adopting and accelerating its own internal use of generative AI tools (dubbed “Elsa”) to augment its review and regulatory activities.

Globally, regulators such as the European Medicines Agency and the International Council for Harmonisation are developing similar AI frameworks, signaling a move toward multi-jurisdictional alignment. Pharmaceutical sponsors should closely monitor these developments and anticipate how evolving regulatory expectations may affect AI integration strategies.

CONCLUSION

AI is revolutionizing how care is delivered, managed, and experienced across the healthcare space, including in connection with the development of medicines. But, with great potential comes great responsibility. Healthcare providers, institutions, and regulators should proactively address the ethical, legal, and regulatory dimensions of AI to ensure that patient rights and public trust remain protected and central.

HOW WE CAN HELP

For questions about how these developments may impact your organization, please reach out to the Morgan Lewis healthcare team.

For more AI in healthcare news, subscribe to Health Law Scan and explore all articles in our AI in Healthcare series.