AstraZeneca Acquires 'Pathology ChatGPT' Modella AI Six Months After Strategic Collaboration

On January 13 local time, during the J.P. Morgan Healthcare Conference, AstraZeneca announced the official acquisition of Modella AI, a Boston-based biomedical artificial intelligence company. The financial terms of the transaction were not disclosed. Following the completion of the acquisition, Modella AI’s generative and agentic AI platforms will be integrated into AstraZeneca’s oncology drug discovery and development framework to accelerate clinical development.

Jorge Reis-Filho, Head of Scientific Innovation AI at AstraZeneca, stated, “The acquisition of Modella AI will bring AstraZeneca cutting-edge pathology foundation models and AI agents, which will support AstraZeneca in continuing to develop targeted therapies and diagnostic tools within its oncology portfolio.”

Furthermore, AstraZeneca CFO Aradhana Sarin stated at the conference that artificial intelligence has been fully integrated into all aspects of AstraZeneca’s operations. From drug discovery and development to commercial operations and healthcare services, AI is helping the company drive outcomes, enhance efficiency and productivity, and accelerate innovation with measurable impact.

In fact, this acquisition builds upon the multi-year partnership agreement announced by both parties in July 2025.Under the agreement, Modella AI will provide AstraZeneca with access to its multimodal AI foundation models to accelerate the clinical development of AstraZeneca’s global oncology portfolio.

Modella AI was founded in 2024 as a spin-off from the Mahmood Lab at Harvard Medical School and Brigham and Women’s Hospital, dedicated to developing generative and agentic AI to advance healthcare, with a particular focus on AI pathology.

The company’s product portfolio primarily consists of two components: the multimodal generative AI model PathChat and the generative AI-powered pathology diagnostic assistant PathChat DX.

PathChat: A Multimodal Generative AI ModelFeatured in the prestigious international journal Nature. The model achieved an accuracy of up to 89.5% when providing useful clinical context, once causing a sensation in the industry; Modella AI has also been dubbed the "ChatGPT of pathology." Building on PathChat, Modella AI has further launched an upgraded version, PathChat 2, which optimizes multimodal generation capabilities, enabling the processing of more complex pathological data and supporting biomarker discovery and personalized treatment planning in clinical trials.

Generative AI Pathology Diagnostic Assistant PathChat DXIt has received FDA Breakthrough Device designation and is the world’s first generative AI–based assistive diagnostic tool for human pathology. In addition, Modella AI offers an AI agent named Judith for automated AI model development. This agent enables functionalities such as cell and tissue segmentation and biomarker identification, supporting gigapixel-level pathological image analysis. It helps researchers rapidly extract key information, thereby accelerating drug development and research into disease mechanisms.

In addition to this acquisition, AstraZeneca has been actively expanding its AI footprint over the past year. In April 2025, AstraZeneca entered into a tripartite collaboration with Tempus AI and Pathos, securing a $200 million data license to build multimodal AI oncology models. In June 2025, AstraZeneca partnered with CSPC Pharmaceutical Group to discover and develop novel oral small-molecule drug candidates, leveraging CSPC’s highly efficient dual-engine AI-driven drug discovery platform; the total value of this collaboration reached up to $5.33 billion. In October 2025, AstraZeneca collaborated with Algen Biotechnologies to leverage their CRISPR-driven functional genomics platform for identifying novel immune-oncology drug targets, with the total deal value amounting to up to $555 million.

However, this acquisition differs from previous collaborations. In the pharmaceutical industry, prior partnerships between large pharmaceutical companies and AI firms have primarily taken the form of strategic alliances, technology licensing, or joint R&D. By directly acquiring Modella AI, AstraZeneca has broken with this traditional model, integrating the AI company into its own R&D system. This marks a shift in the role of artificial intelligence within drug discovery, elevating it from an “external tool” to “core infrastructure.”

At the 2026 J.P. Morgan Healthcare Conference, Eli Lilly also announced collaborations in the field of AI.On January 12 local time, NVIDIA and Eli Lilly announced a $1 billion investment to establish a new joint research laboratory in the San Francisco Bay Area over the next five years, aiming to accelerate AI-driven drug discovery. The laboratory will leverage NVIDIA’s latest Vera Rubin architecture AI chips to support cutting-edge biomedical research, with the goal of reshaping drug development processes in the era of artificial intelligence. It will also integrate Eli Lilly’s expertise in drug discovery and clinical research. Related operations at the laboratory are scheduled to officially commence in early 2026.

Meanwhile, NVIDIA also disclosed that it is collaborating with Thermo Fisher Scientific to build highly automated “autonomous laboratory” infrastructure.

Not only are AstraZeneca, Eli Lilly, and Thermo Fisher Scientific racing to secure a foothold in the AI sector, but in recent years, pharmaceutical giants such as Roche, Novartis, Gilead Sciences, and Sanofi have also heavily invested in AI-driven drug discovery.According to data published by Nature in December 2025, AI-driven drug discovery is emerging as a rapidly growing segment in licensing deals in 2025. Since 2017, there have been 513 such transactions globally, with 120 occurring in the first ten months of 2025 alone, accounting for 23% of the total.

The frequent emergence of business development (BD) deals in this field is driven by the significant application value that AI offers to drug discovery and development.

Specifically, for AI companies,Collaboration with large pharmaceutical companies presents an opportunity to access essential resources, accelerate drug development processes, and expand business scope. AI-driven pharmaceutical companies often possess deep expertise in specific technological areas or disease fields. If an AI pharmaceutical company can offer unique technologies that address the pain points of traditional pharmaceutical firms and help them understand the technical advantages of its AI-driven platform, it is likely to be favored by these industry giants.

For large pharmaceutical companies,There is widespread optimism regarding the application of AI in drug development, with AI regarded as a key tool for enhancing efficiency, reducing costs, and accelerating time-to-market for new drugs. Meanwhile, although large pharmaceutical companies possess substantial funding, extensive data resources, and rich experience in drug R&D, they generally do not develop all tools in-house; in most cases, they prefer to leverage tools developed by specialized firms.

However, as the industry continues to advance, we must also confront the current realities and pain points of AI-driven drug discovery: the application of AI in this field is transitioning from “proof of concept” to “scaled implementation.” While AI technologies have demonstrated significant advantages in areas such as protein structure prediction (e.g., AlphaFold), virtual screening, and molecular generation, their actual effectiveness in clinical translation still requires further validation through additional data. Currently, AI platforms are primarily focused on the early discovery stage, whereas the majority of costs and risks in drug development are concentrated in the clinical trial phase. Therefore, whether AI technology can truly shorten the overall R&D cycle and reduce failure rates remains to be tested over the long term by clinical outcomes. The industry generally expects that as the first batch of AI-designed drugs gains regulatory approval and enters the market in the coming years, the commercial value of this technology will be more clearly validated, thereby driving increased upfront investment.

References:
“Pathology’s ChatGPT” Partners with AstraZeneca! Published in Nature, Achieving 89.5% Accuracy!

“Three Major Trends in Biopharmaceutical Licensing Deals in 2025: Chinese Strength, Multispecific Antibodies, and AI-Driven Drug Discovery”