Vivodyne’s HIVE robotic labs grow 20 types of human tissue to supply causal data for AI drug models.

Vivodyne opened the world’s largest human data center near San Francisco, using its HIVE robotic labs to grow 20 types of human tissue and test drugs on them. The system autonomously doses and monitors the tissue, generating causal biological data that today’s AI models lack.

The AI drug-discovery industry has produced few validated drugs despite claims from OpenAI’s Sam Altman, Google DeepMind’s Demis Hassabis, and Anthropic’s Dario Amodei. AlphaFold, a Nobel-prize-winning tool, has yet to yield a new drug, and Isomorphic Labs expects its first trials by the end of this year.

For builders and operators, the takeaway is that AI models trained on animal or static cell data cannot predict human responses. Vivodyne claims its liver tissue matches human toxicity trials with 94% accuracy, airway tissue at 96%, and bone marrow at 100% concordance in chemotherapy tests.

Watch for Vivodyne’s tissue data to enter clinical trials and for Isomorphic Labs to start its own trials by the end of this year. Vivodyne, backed by Khosla Ventures, raised just under $80 million and claims its labs have twice the throughput of US animal testing.

What matters

• Vivodyne opened a human data center with robotic labs that test drugs on living tissue.
• AI drug discovery models lack causal data from human tissue, which limits their real-world use.
• Watch for Vivodyne’s clinical trial performance and whether its tissue data leads to approved drugs.

Why it matters

Watch for Vivodyne’s clinical trial performance and whether its tissue data leads to approved drugs.

This GenAI News article was prepared in original wording using reporting and materials published by TechCrunch AI. Source reference: https://techcrunch.com/2026/08/19/ai-isnt-close-to-curing-cancer-this-startup-says-it-knows-what-it-will-take/.

Drafted by the GenAI News review pipeline.