New York-based Pathos AI has raised $237 million in Series E funding at a $3 billion post-money valuation, as it expands an AI-powered drug development platform and clinical-stage oncology pipeline.
The financing was led by pharmaceutical company Bristol Myers Squibb, with participation from existing investors.
Unlike biotechnology startups focused exclusively on selling AI software to pharmaceutical companies, Pathos is using its technology to develop and manage its own portfolio of drug candidates.
The company believes AI can improve decisions about which treatments to pursue, how to design clinical trials and which patients are most likely to benefit.
How Pathos Uses AI in Drug Development
Traditional pharmaceutical development involves numerous decisions made with incomplete information.
Researchers must identify promising compounds, determine which diseases to target, design clinical trials and decide whether to continue investing after receiving new results.
A drug candidate can fail because the underlying biology is incorrect, the treatment is insufficiently effective or the trial enrolls patients who are unlikely to respond.
Pathos is building Foundry, a platform that uses thousands of specialized AI agents to analyze information across the drug development process.
The system combines clinical trial data, genomics, biological samples and information about competing therapies.
According to Pathos, Foundry continuously evaluates the available evidence and estimates the probability of success for different development strategies.
Its intended applications include identifying patient populations, evaluating treatment mechanisms and informing modifications to clinical development programs.
The objective is not simply discovering new molecules.
Pathos wants AI to influence the full process of selecting, developing and testing medicines.
Who Founded Pathos AI?
Pathos was founded in 2022 by Eric Lefkofsky and Ryan Fukushima.
Lefkofsky also founded healthcare data company Tempus, which uses genomic and clinical information to support precision medicine.
Iker Huerga joined Pathos as CEO in 2025.
Huerga previously worked in oncology research and development at AstraZeneca and held leadership roles at Tempus.
The company's team combines pharmaceutical development experience with expertise in AI and clinical data.
Pathos' Oncology Drug Pipeline
Pathos reports four active clinical-stage assets.
Its lead priorities include JSKN016, a TROP2/HER3-targeting bispecific antibody-drug conjugate designed to deliver a therapeutic payload to selected cancer cells.
The company plans to advance the program toward registrational studies.
Other assets include pocenbrodib, being investigated in prostate cancer and multiple myeloma; AZD4241, an estrogen receptor-targeting PROTAC program partnered with AstraZeneca; and DO-2, a MET kinase inhibitor.
These approaches target different biological mechanisms.
The diversification gives Pathos several potential opportunities for clinical success, but it also increases development costs and operational complexity.
The company has not demonstrated that its AI platform can reliably predict regulatory approval or clinical success across its portfolio.
Expanding Into Neurological Diseases
Alongside the funding, Pathos announced collaborations with Bristol Myers Squibb and Tempus involving neurological conditions, including Alzheimer's disease.
Under the agreement, Tempus will develop a de-identified longitudinal dataset spanning hundreds of thousands of patients with neurological conditions.
Pathos will license the data through a five-year agreement worth $55 million.
Bristol Myers Squibb will gain access to Pathos Foundry for certain research and development applications.
This expansion is significant because neurological drug development presents challenges different from oncology.
Disease progression can be difficult to measure, patient populations are heterogeneous and clinical outcomes may require years of observation.
Pathos is betting that richer longitudinal datasets can improve decisions about disease biology and clinical trials.
Competition and the Fundamental Risk
Pathos operates in a market that includes AI drug discovery companies such as Recursion, Insilico Medicine and other biotechnology developers using computational methods.
Its differentiation is the combination of AI-driven portfolio decisions with ownership of clinical-stage assets.
That model potentially captures more value if a drug succeeds, but it also exposes investors to the full cost and risk of pharmaceutical development.
Clinical trials remain expensive, lengthy and uncertain.
An AI system may identify correlations in existing patient data without discovering biological relationships that generalize to new populations.
Models also risk producing overly optimistic probability estimates if historical data is incomplete or biased.
Pathos' $3 billion valuation reflects confidence in AI-assisted drug development. The decisive evidence will come from clinical outcomes, not from the number of agents operating inside Foundry or the speed at which the company expands its pipeline.