AI has long been touted as the key to curing cancer, but a biotech startup argues the industry is missing a critical ingredient: the right data. Vivodyne, a company spun out of the University of Pennsylvania, claims it has built a solution to bridge this gap and accelerate the path to real breakthroughs.
At the heart of Vivodyne's approach is HIVE, a network of modular robotic labs that grow 20 types of human tissue. These labs autonomously dose and monitor the tissues, generating causal biological data that current AI models lack. Today, much of the training data for AI in drug discovery comes from animal testing or studies of isolated cells and proteins—not living human tissue. Vivodyne aims to change that.
“Absent human testing, what are these [AI] models going to do?” asks Andrei Georgescu, Vivodyne's CEO and co-founder. “They’re going to cure cancer in mice.”
The skepticism is growing. Even Dario Amodei, CEO of Anthropic, recently wrote that claims about AI curing cancer have become “more cliche than credible.” He added, “The thing that will work is actually curing cancer.” Ironically, Amodei himself has floated such ideas in the past. Sam Altman has repeatedly cited curing cancer as a justification for OpenAI's push toward artificial general intelligence (AGI) and massive compute buildouts. Google DeepMind's Demis Hassabis also said last year that AI could potentially cure all disease within a decade.
Yet, real-world progress remains modest. A few AI-designed drugs have entered human trials, with one reaching Phase III—the final stage before regulatory approval. But the roadblocks are not purely computational. As of 2026, no AI-discovered drug has been approved, and the challenges lie in translating predictions into safe, effective treatments for humans.
Nobel Prize-winning AlphaFold marked a major advance in understanding protein structures, but it has yet to produce a new drug. Isomorphic Labs, founded to build on AlphaFold's work, originally planned its first trials for 2025 but now expects them by the end of this year. In February, the company noted that true drug discovery will require “highly accurate predictive models, across an expansive range of biochemical properties and interactions.”
Georgescu calls for “a sanity check” on AI's promises, noting that existing models lack the data to capture human biology's complexity. This is a pressing issue for the pharmaceutical industry, where 90% of drugs that succeed in animal trials fail to gain regulatory approval for human use.
Vivodyne's strategy is distinct. Founded in 2021 after Georgescu completed his PhD in bioengineering at Penn, the company claims its tissues closely mimic human organ behavior. For instance, its liver cells show 94% predictive accuracy in toxicity trials, its airway tissue matches human responses 96% of the time, and its bone marrow has achieved 100% concordance in tests of 20 chemotherapy drugs.
Last week, Vivodyne opened what it calls the world's largest “human data center” near San Francisco, backed by nearly $80 million raised across two rounds led by Khosla Ventures. Georgescu says the facility is already generating twice the throughput of all animal trials conducted in the United States, offering a scalable and ethical alternative.
The biolab of the future? Credit: TechCrunch/Tim Fernholz
The goal is to accelerate drug discovery by providing AI systems with the high-quality, causal data they need to make reliable predictions. While challenges remain—such as scaling regulatory acceptance and ensuring long-term safety—Vivodyne's approach may offer a more grounded path to turning AI's potential into tangible medical breakthroughs. As the industry moves toward 2027, the focus is shifting from hype to validation, and data may be the key differentiator.
via TechCrunch AI