As the landscape of drug discovery becomes increasingly complex, the generation and ownership of data in partnerships between Big Pharma and AI-driven biotechs are creating new headaches for dealmakers. Today, the ones and zeros required to train AI models for better drug target selection have become a highly valuable currency in biopharma transactions.
When structuring major agreements, AI drug hunters face intricate negotiations to ensure their platforms can continue to learn and advance. Conversely, pharmaceutical giants must adapt to this new normal if they want to capitalize on next-generation technology.
The Shift in Deal Dynamics
The complexity of these modern agreements is undeniable. “I just turn on my ‘out of office’ when I hear a data deal is coming,” joked Chad Diehl, legal team lead for licensing and acquisitions at Astellas Pharma, during a recent industry panel.
Traditionally, pharma companies demanded strict control over any data developed during a partnership. According to Deepa Talpade, head of business development and licensing for oncology at Bayer, even granular details—such as whose chemistry lab is used or where data is physically generated—are now pivotal decision points at the negotiating table.
In the past, biotech platform companies would typically hand over most of the data generated in a licensing deal, keeping partnership programs strictly siloed. While they used general learnings to improve their platforms, the raw data itself wasn’t fed back into their algorithms for use on other projects.
Rachel Lane, senior vice president of business development and operations at Xaira Therapeutics, noted that this outdated model is no longer viable for AI-focused agreements. Because AI models require constant feeding to improve, companies running complex labs view these deals primarily as a mechanism to access crucial new datasets they wouldn’t otherwise possess.
Big Pharma Stays in the Game
Despite these legal and logistical hurdles, pharmaceutical companies are actively pursuing these partnerships. Merck recently signed an agreement with Protillion Biosciences worth $510 million upfront, while Eli Lilly, Bristol Myers Squibb, and Incyte forged similar deals with AI companies earlier this year.
Eli Lilly has also built massive internal AI capabilities, establishing TuneLab to grant select biotechs access to the pharma giant’s AI suite in exchange for data sharing. This initiative highlights that the thirst for high-quality data points goes both ways.
Navigating the New Landscape
Biotechs using AI to discover new drugs must understand the leverage their data provides, while still recognizing that partnerships are essential to proving their concepts and advancing their work. When structured correctly, the mutual benefits are immense. “One plus one equals five in the context of the deal,” Lane explained.
To navigate these complex data issues, Maha Radhakrishnan, an executive partner at Sofinnova Investments, emphasized the need for specialized expertise. She advised bringing in patent attorneys and legal teams from the very beginning to establish clear boundaries and ensure total clarity around data ownership before any deal is signed.
Source: Biospace | Jul 22, 2026
