AstraZeneca is making a deeper bet on artificial intelligence as it looks to speed up and sharpen its oncology research. The pharmaceutical giant has agreed to acquire Boston-based Modella AI, bringing the company’s technology, data, and staff directly into its research organization. Financial terms of the deal were not disclosed.
The move reflects a broader shift in drug development, where the question is no longer whether AI can help, but how closely it needs to be embedded in daily research and clinical decision-making. For AstraZeneca, the answer appears to be full integration.
From partnership to ownership
AstraZeneca and Modella AI have worked together for several years, using AI to analyze pathology data such as biopsy images and connect those findings with clinical information. The goal has been to make pathology more quantitative, helping researchers identify biomarkers and better understand how patients might respond to specific treatments.
Over time, that collaboration revealed the limits of an external partnership. Speaking at the J.P. Morgan Healthcare Conference, AstraZeneca Chief Financial Officer Aradhana Sarin said the company wanted more of its data and AI capabilities under one roof as oncology research becomes increasingly complex and time-sensitive.
By acquiring Modella, AstraZeneca is moving its AI tools from a supporting role into the core of its oncology research and development work. Modella said its foundation models and AI agents will now be integrated across clinical development and biomarker discovery efforts.
Why AI ownership matters in drug development
The deal highlights why ownership is becoming more important for large pharmaceutical companies. Developing drugs today generates vast amounts of data, and using AI effectively depends on consistent access to that data, tight integration with existing workflows, and the ability to adapt tools as research needs evolve.
Gabi Raia, Modella AI’s chief commercial officer, said joining AstraZeneca will allow the company’s technology to be deployed across global trials and clinical settings, rather than being limited to pilot projects or isolated studies.
Rather than relying on external vendors and fixed product roadmaps, AstraZeneca will now directly shape how Modella’s models are developed and applied, an advantage in a highly regulated environment where transparency and validation are critical.
Improving trial decisions and patient selection
One of the most immediate applications of AI for AstraZeneca is in clinical trials. By improving how pathology and clinical data are analyzed together, the company hopes to make better decisions about trial design and patient selection.
Matching the right patients to the right trials can improve outcomes and reduce the risk of costly delays or failed studies. While advanced algorithms play a role, executives emphasize that the real gains come from reliable data pipelines and tools that fit naturally into researchers’ day-to-day work.
Sarin said the acquisition would strengthen AstraZeneca’s capabilities in quantitative pathology and biomarker discovery, helping translate research insights into practical decisions more quickly.
Talent moves inside the lab
The acquisition also brings Modella’s data scientists and AI specialists into AstraZeneca’s internal teams. This reflects a growing trend in the industry, where AI talent is increasingly viewed as essential to core research rather than as an external service.
AstraZeneca said this marks the first time a major pharmaceutical company has acquired an AI firm outright, even as collaborations between drugmakers and technology companies continue to multiply.
At the same healthcare conference, other major deals underscored that interest. Nvidia and Eli Lilly, for example, announced a $1 billion partnership to build a new research lab using Nvidia’s AI chips. The contrast highlights two approaches: partnerships that accelerate experimentation, and acquisitions that signal a longer-term commitment to internal capability.
A long-term bet on embedded AI
Sarin described the earlier AstraZeneca–Modella collaboration as a “test drive” that ultimately showed the company wanted deeper control over the technology. The broader ambition is to support the development of highly targeted biomarkers and, in turn, more precise cancer therapies.
Looking ahead, AstraZeneca expects 2026 to be a busy year, with several late-stage trial results anticipated across multiple therapy areas. The company is also working toward a goal of $80 billion in annual revenue by 2030.
Whether in-house AI delivers on that promise will depend on execution. Integrating AI into drug development is complex and often slow. Still, AstraZeneca’s move makes its strategy clear: the company sees lasting value not in buying AI as a service, but in embedding it deeply into how new medicines are discovered and tested.
