How AI And Mini-Organs Could Replace Testing Drugs On Animals

They’re also central to the future of pharmaceutical testing—and could one day make lab animals obsolete.

Since 1937, animal testing has been a legal requirement, prompted by a deadly incident involving a contaminated antibiotic that caused over 100 deaths. Nearly 90 years later, despite rigorous animal trials, drugs continue to be withdrawn from the market due to unforeseen toxic effects in humans. Today, a growing coalition of policymakers, scientists, and innovators is championing advanced alternatives to animal testing—methods that could enhance drug safety, reduce development costs, and accelerate the path to clinical trials.

In 2022, researchers tested 27 drugs previously deemed safe in animal studies—some of which had later been pulled from the market due to human toxicity. Using a cutting-edge technology known as “organ-on-a-chip,” they were able to accurately identify the harmful compounds. These chips, akin to organoids, contain living human cells embedded in microchips that mimic the function of real organs. The liver-on-a-chip model proved particularly effective at detecting toxicity. According to the study, widespread adoption of such technology could save the pharmaceutical industry more than $3 billion annually by reducing costly late-stage drug failures.

Beyond safety, cost is a major driver for change. Bringing a single drug to market now exceeds $2 billion on average, with the global industry investing close to $300 billion each year in R&D. Yet, more than 90% of experimental drugs fail—often because results from animal trials don’t translate to humans. This inefficiency contributes heavily to the sky-high prices of approved medications.

Animal testing is a significant bottleneck. Despite passing preclinical animal trials, most drugs fail in human trials, highlighting the biological mismatch between species. As Ali Afshar, CEO of London-based Mytos, puts it: “Everything that reaches clinical trials has already succeeded in animals. So clearly, we’re not getting accurate predictions.”

Mytos is developing an automated system to grow human cell cultures in petri dishes, offering a more consistent and human-relevant platform for drug testing. Traditional manual cell culturing often leads to variability, undermining reproducibility. Automation not only speeds up the process but improves data reliability. Since its founding in 2016, Mytos has raised nearly $29 million and now supplies its cell models to pharmaceutical companies, particularly for diseases where animal models fall short.

Organoids and advanced cell cultures are among the alternatives the FDA is now encouraging—especially for monoclonal antibodies, a class of drugs used to treat conditions ranging from cancer to autoimmune diseases and viral infections like COVID-19. These therapies often don’t work in mice and require testing in primates, which can cost tens of thousands of dollars per animal. Even then, results frequently fail to predict human outcomes. In April, the FDA issued new guidance suggesting developers use human-based alternatives to demonstrate safety for these drugs, marking a shift toward human-relevant data.

This shift was made possible by the FDA Modernization Act 2.0, signed into law by President Joe Biden in 2022 after unanimous Senate approval. The law removes the mandatory requirement for animal testing when alternative methods—such as computer modeling or organ-on-a-chip systems—provide sufficient safety evidence. It also waives the requirement for biosimilars, drugs similar to already-approved biologics. Support for reducing animal testing has persisted into the second Trump administration, with the National Institutes of Health launching a coordinated effort to fund and promote non-animal research technologies.

Still, every new method comes with limitations. Julie Frearson, chief scientific officer at Charles River Laboratories, notes that organoids, while powerful, offer only a narrow view. They can show how a drug affects a specific tissue—like heart or liver tissue—but can’t reveal systemic effects across the entire body. Additionally, organoids are short-lived, making it difficult to study long-term drug impacts.

San Francisco’s Gordian Biotechnology is tackling this challenge with a technique called mosaic screening. By introducing multiple gene therapies into a single cell within a living animal, they can observe how each therapy behaves over time within a whole-body system. This allows for long-term, systemic data while drastically reducing the number of animals needed. The $170 million company uses larger, more human-like animals such as horses—whose aging diseases more closely mirror those in humans—rather than mice. They’re currently developing gene therapies for osteoarthritis and fatty liver disease.

“The fundamental problem every biotech faces is that animals aren’t humans, and neither are organoids or isolated cells,” said Gordian CEO Francisco LePort. “You won’t truly know how a drug works until you test it in a human.”

To bridge this gap, companies are building more sophisticated organoids. Vivodyne, a San Francisco startup, recently secured $40 million in Series A funding to develop larger, more functional organoids that better mimic real organs—including some capable of circulating blood. CEO Andrei Georgescu believes these advanced models will allow earlier and more accurate predictions of drug efficacy. “Right now,” he said, “we lack enough reliable data to draw strong correlations. That’s the real bottleneck.”

Artificial intelligence is also emerging as a key player. The FDA Modernization 2.0 Act and recent NIH guidelines both recognize AI as a promising alternative to animal testing. While AI has deepened our understanding of human biology, experts caution that we’re still far from fully simulating human drug responses.

“We probably understand only 10% to 15% of biology’s core principles,” said Najat Khan, chief R&D officer at Recursion, a company using AI for drug discovery with several cancer therapies in human trials.

Parallel Bio, a Cambridge, Massachusetts-based biotech with $30 million in funding, combines AI with patient-derived organoid lymph nodes to model immune responses. By testing drug candidates across diverse demographic samples and feeding the data into AI algorithms, they aim to predict safety and effectiveness more accurately for pharmaceutical partners like vaccine developer Centivax. CEO Robert DiFazio says their mission is clear: “Move drug discovery away from animal models and into systems that better reflect human biology.”

While DiFazio acknowledges that animal testing may never be fully eliminated, the momentum toward alternatives is strong and broadly supported.

“This is one of the few issues today with solid bipartisan backing,” he said. “People understand—this is why drugs are so expensive, and why developing new medicines is so slow and risky.”

More from Forbes

ForbesVolunteers Track The $10 Billion Blow To American ScienceBy John DrakeForbesStart-Ups Eye The Opportunity In AI Software Testing MarketBy David ProsserForbesVibe Coding Turned This Swedish AI Unicorn Into The Fastest Growing Software Startup EverBy Iain Martin

The above is the detailed content of How AI And Mini-Organs Could Replace Testing Drugs On Animals. For more information, please follow other related articles on the PHP Chinese website!