Life science companies are rapidly embracing the promise of using AI to accelerate discovery and development. That enthusiasm may be tempered as companies that in-license a product developed with AI consider the implications of such use on their ability to obtain regulatory approval. We highlight below a few manners in which AI is currently being used in discovery and development, and suggest what licensees should proactively address in their licenses in light of such uses.

AI in action

• AI is being used to identify novel molecules and targets or in some cases, embedded directly into regulated products.
• AI is also being used to design clinical trials, generate and evaluate data, including safety signals and target-identification data, all of which is used in regulatory submissions.
• As AI becomes embedded throughout the development lifecycle, companies must ensure adequate transparency into model inputs, assumptions, and limitations so that AI generated outputs can withstand regulatory scrutiny and meaningfully support development decisions.

License considerations

• Licensors should be required to provide all necessary validation materials and sufficient visibility into the scientific inputs and assumptions underlying any algorithms used to verify that the outputs can meaningfully support progression into preclinical and clinical development.
• Licensees should obligate licensors to secure any rights and sublicenses from the AI platform provider to ensure they can use and submit such materials, assumptions, and algorithms to regulatory authorities and accurately respond to any inquiries from the regulatory authorities.
• Licensees will want to consider the impact of not obtaining such materials, rights, and sublicenses on their diligence obligations (e.g., what happens if licensors cannot provide any required AI-related documentation to support regulatory approval and the licensee has to restart clinical development).
• Licensees will also want to consider the impact of any failures or inaccuracies in any such materials, including in any AI-generated information with specific warranties and indemnities, to appropriately allocate risks arising from licensors use of the AI tools.

AI has brought significant advantages to the table in life sciences product development, offering new speed, scale, and insights at all stages of the development program. Robust diligence and smart contracting will allow licensees to capture the upside while keeping risk in check.