Recursion and Genentech say they have validated their first drug target discovered through artificial intelligence — a novel neuroscience target found by mapping the biology of lab-grown brain cells at massive scale.
The two companies have collaborated since 2021 in a partnership worth up to $12 billion in potential milestones. As of August 2026, Genentech has paid Recursion about $216 million, including a recent $3 million milestone tied to this target. Each program can earn Recursion up to $300 million in milestones plus tiered royalties. The specific target and disease were not disclosed.
How the AI found it
Recursion built the finding on its Data Factory: a whole-genome CRISPR knockout map made from neurons derived from induced pluripotent stem cells. The scale is striking — more than 1 trillion lab-grown neuronal cells (roughly a dozen brains’ worth) and over 46 million cellular images analyzed on the company’s BioHive-2 supercomputer. From initiation to a validated target package took just 15 months.
Why it matters
“Finding new targets in neuroscience has historically been challenging,” said Recursion CEO Najat Khan, adding that the milestone “highlights our ability to uncover novel biology in areas where conventional approaches struggled.” The broader collaboration spans 40 programs and six whole-genome “phenomaps,” concentrated in neuroscience plus one gastrointestinal-oncology indication. A validated target is an early step — years of drug development still lie ahead — but it is a concrete result from the much-hyped promise of AI-driven discovery.