Human Brains Kept Alive For Drug Tests

Group of scientists in a laboratory working together on experiments

A Connecticut biotech startup is keeping freshly extracted human brains on life support to test experimental drugs — and it just raised over $42 million to expand the practice.

Story Snapshot

Yale spinout Bexorg extracts donated human brains and places them on perfusion machines to restore cellular activity, then uses the brains as drug-testing platforms.
Brain drug trials currently fail 95–99% of the time, and Bexorg argues testing on real human tissue — rather than rodents — can dramatically improve those odds.
Bexorg confirms the research brains never exhibit electrical activity associated with thought or sensation and cannot be brought back to consciousness.
The company has raised $42.5 million in funding and is positioning the platform as a replacement for animal testing in central nervous system drug development.

Brains on Machines: What Bexorg Is Actually Doing

Bexorg, a New Haven, Connecticut startup spun out of Yale University, has developed what it calls a “whole human brain drug discovery platform.” The company receives donated human brains, places them on perfusion support systems that restore blood flow and molecular activity, and then administers experimental drug compounds directly through the brain’s intact vascular system and blood-brain barrier. Researchers can biopsy the tissue to measure how long a drug stays in cells, whether it hits its intended molecular target, and any early signs of side effects.

Bexorg describes its platform as capable of delivering “translational data impossible to obtain from animal or in vitro” testing models. The company markets itself as “the first-ever AI and whole-human brain drug discovery platform built to end central nervous system (CNS) clinical failure.” Funding has followed: the company has raised $42.5 million to scale operations, with industry observers describing the platform as potentially transformative for a drug development sector plagued by catastrophic failure rates.

The 95–99% Failure Rate Driving Demand

The core argument behind Bexorg’s approach rests on a well-documented crisis in brain drug development. According to Yale Ventures, brain drug trials currently fail 95–99% of the time. Researchers and investors widely attribute this failure rate to the fundamental mismatch between rodent biology — the standard preclinical testing model — and the human brain. Bexorg argues that testing compounds in a real human brain environment before launching costly clinical trials gives pharmaceutical companies far more reliable data on how a drug will actually perform in patients.

The platform allows researchers to administer therapeutic compounds and monitor brain responses in real time across multiple molecular layers. By catching failures earlier — in a human tissue environment rather than in expensive late-stage human trials — the approach could theoretically save billions of dollars and years of wasted development time. That promise has attracted serious investment and attention from pharmaceutical companies looking for better predictive tools before committing to full clinical programs.

Consciousness Questions and Ethical Boundaries

The practice raises immediate and legitimate questions that any thoughtful person should ask. Bexorg directly addresses the most pressing concern in its own materials: the research brains “never exhibit any of the electrical activity necessary for thought or physical sensation,” and the perfusion process “can never bring the research brain back to consciousness.” The company draws a clear line between restoring cellular and molecular activity — which it does — and restoring any form of awareness or personhood — which it says is impossible under its protocol.

Even so, the ethical framework surrounding this technology remains unsettled. Critics and ethicists point out that consent protocols for whole-brain donation differ meaningfully from standard organ or tissue donation, and questions about reproducibility and whether cellular signals reliably predict clinical outcomes in living patients have not been fully answered. The science may be genuinely promising — the failure rate problem is real and the human-relevance argument is sound — but the broader validation of whether these brain-derived signals actually translate into better drug approvals remains an open question that independent researchers have yet to fully resolve.