Loading page…
Tempus will assemble patient data for Pathos’s AI platform. Bristol Myers gains three-year exclusive neuroscience rights, while better trials remain the goal.
Listen to this story
Pathos AI, Tempus AI and Bristol Myers Squibb announced a collaboration on October 23, 2026, to apply AI and drug-development expertise to Alzheimer’s and other neurological diseases. Pathos will pay $55 million over five years to license a planned longitudinal dataset covering hundreds of thousands of patients, then use its Foundry platform to support disease research and clinical-trial planning. The dataset has not yet been built, and the collaboration is intended to inform drug development—not provide a treatment or establish improved neurological outcomes.
Tempus will assemble the neurological dataset using infrastructure and expertise it developed in oncology.
Bristol Myers will have three-year exclusive Foundry access for neuroscience programs, but only non-exclusive strategic access across its broader portfolio.
Pathos says Foundry supports adaptive trial design in oncology; that company claim is not evidence of improved outcomes in this neurological collaboration.
Pathos AI is bringing patient records, AI analysis and pharmaceutical development expertise into a collaboration aimed at Alzheimer’s disease and other neurological conditions. Announced October 23, 2026, the agreements with Tempus AI and Bristol Myers Squibb pair a planned large-scale dataset with tools intended to guide research and clinical-trial design—not a newly developed treatment.
Tempus will build a de-identified dataset covering hundreds of thousands of patients with neurological conditions. It is designed to follow patient outcomes over time, rather than provide only a single snapshot of disease. Pathos will license that dataset through a five-year, $55 million agreement.
The dataset is still to be created. Tempus plans to use infrastructure and expertise it developed in oncology—the study and treatment of cancer—to assemble this neurological resource. Its assigned role is to supply curated patient information that can support research into disease biology and progression.
That establishes a specific division of work: Tempus builds the data resource; Pathos applies its Foundry AI platform; Bristol Myers contributes neuroscience drug-development expertise. The companies describe those pieces as complementary. The collaboration’s starting point is understanding neurological disease and using that understanding to advance potential therapies.
Neurological diseases demand more than any one company can bring on its own: the best longitudinal data, AI that can reason over it at scale, and the drug development expertise to turn insight into medicines.
Iker Huerga, CEO of Pathos AI
Pathos says Foundry will analyze the dataset at scale, generate and test hypotheses about disease biology, and inform more targeted, adaptive clinical trials. Those are linked but distinct tasks: investigating how a disease works, then using the findings to help decide how a potential treatment should be studied.
Bristol Myers’s stated goals make the intended clinical use more concrete. It wants deeper insight into disease progression, help identifying patients most likely to benefit from potential therapies, and more precise trial designs. The emphasis is on decisions within drug development, rather than an AI service giving treatment advice directly to patients.
Pathos says it built and validated this approach in oncology, where Foundry helps guide its own clinical-stage drug pipeline. The company describes adapting trial designs in real time to improve their projected probability of success. That is a company claim about its existing approach—not a demonstrated improvement in neurological treatment outcomes from the new collaboration.
Bristol Myers will receive three-year exclusive rights to use Foundry for its neuroscience development programs. Across its broader portfolio, its strategic access to the platform will be non-exclusive. The announcement therefore gives the drugmaker different access terms depending on the area of research, rather than blanket exclusivity over Foundry.
The data license and platform rights also run on different clocks. Pathos’s agreement to license the Tempus dataset lasts five years; Bristol Myers’s neuroscience exclusivity lasts three. These are separate commitments within the collaboration: one secures the information Pathos intends to analyze, while the other defines a partner’s rights to use the AI platform.
Story updates
Pathos AI has fresh capital to develop its own cancer drugs while bringing its drug-development software into neuroscience. The company said on October 23, 2026, that it closed a $237 million Series E led by Bristol Myers Squibb, valuing it at approximately $3 billion after the investment. Linked partnerships extend its work to Alzheimer’s disease…
Pathos AI has fresh capital to develop its own cancer drugs while bringing its drug-development software into neuroscience. The company said on October 23, 2026, that it closed a $237 million Series E led by Bristol Myers Squibb, valuing it at approximately $3 billion after the investment. Linked partnerships extend its work to Alzheimer’s disease and other neurological conditions.
Bristol Myers is taking two roles in the arrangement: financing Pathos and gaining access to Foundry, its AI platform. Existing investors also participated in the round. Pathos says the proceeds will support both further platform development and clinical work on a pipeline that now contains four active drug candidates.
A key funding priority is JSKN016, a cancer drug candidate that Pathos intends to advance toward registrational studies. The company says Foundry helped it identify the asset. Its financing announcement also names three other clinical programs that will receive support, making the use of proceeds more specific than a general commitment to expand AI research.
Pocenbrodib: funding will support clinical development in prostate cancer and multiple myeloma.
AZD4241: Pathos will advance this program, which is partnered with AstraZeneca.
DO-2: the financing will also support clinical advancement of this next-generation MET kinase inhibitor.
Pathos describes itself as a company developing drugs, not just supplying software to other developers. Its stated approach connects selecting clinical assets, making portfolio decisions and monitoring trials. The four-program portfolio therefore gives its AI system a direct role in decisions about medicines the company itself is advancing.
Foundry uses thousands of specialized AI agents to bring together information about a drug program, according to Pathos. Those agents analyze clinical data, genomics, biological specimens and competing drug-development efforts. The platform continuously updates estimates of a program’s probability of success rather than treating the initial decision to pursue a drug as fixed.
The intended outputs are practical choices: which drugs to pursue, which patient groups to study and how to design clinical trials. Pathos says the system can help identify biological subgroups, guide changes to trials and rule out high-risk development paths early. These are the company’s descriptions of Foundry’s capabilities, not demonstrated improvements in patient outcomes.
The neuroscience expansion depends on a data resource still to be built. Tempus will assemble de-identified information from hundreds of thousands of patients with neurological conditions. The dataset is intended to follow outcomes over time, using infrastructure and expertise Tempus developed in oncology, the study and treatment of cancer.
Pathos plans to apply Foundry to that resource to generate and test ideas about disease biology and inform more targeted, adaptive clinical trials. Bristol Myers contributes neuroscience research and clinical-development expertise. Its stated goals include understanding disease progression and identifying patients most likely to benefit from potential therapies.
That extends an approach Pathos says it built and validated in oncology into a different disease area. Bristol Myers frames the collaboration as a way to combine AI analysis with scientific expertise, with the hope of improving trial design and advancing treatment options. The neurological work is a research and development plan, not an announcement of a new Alzheimer’s treatment.
Loading discussion...
Join the conversation
Explain whether exclusivity would help or hinder the search for treatments.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.