BioTech

Isomorphic Labs: Hassabis’ Pharma AI Set for 5 Billion Dollar Mega-Round Without a Single Drug Tested in Humans

co-produced by AI05. November 2026, 07:23

Demis Hassabis is raising more money for a company without a single drug tested in humans than most pharma giants spend on research in a year.

Isomorphic Labs, the AI drug discovery company spun out of Google DeepMind in 2021, is in talks to raise around 5 billion dollars (roughly 4.4 billion euros), according to Bloomberg. The valuation is said to be about 50 billion dollars (roughly 44 billion euros). The talks are at an early stage, terms could still change, and the investors expected to back the deal have not been named. In early October, Bloomberg had already reported on talks at a valuation of at least 40 billion dollars, a figure that has apparently since been revised upward.

Two Rounds in Six Months

The pace is remarkable even by AI sector standards. Only in May did Isomorphic Labs close a 2.1 billion dollar Series B, led by Thrive Capital, with Alphabet, GV, CapitalG, MGX, Temasek and the UK Sovereign AI Fund on board, at the time the largest round ever for an AI biotech. Before that, there was a first external round of 600 million dollars in 2025, also led by Thrive. If the new round goes through, Isomorphic will have raised close to 7.7 billion dollars in roughly 18 months, at a valuation that market observers say has at least doubled since May.

Hassabis, who runs Isomorphic alongside his job as DeepMind CEO, won the 2024 Nobel Prize in Chemistry for AlphaFold, the AI model that predicts the three-dimensional structure of proteins. Isomorphic builds on that: its in-house drug design engine IsoDDE is meant to design drug molecules on a computer rather than hunt for them in years of lab work. The stated goal is to cut drug development time from around ten years to a fraction of that and one day “solve all disease.” Its own pipeline focuses on oncology and immunology.

Where the Money Has Come From So Far

Isomorphic’s revenue to date comes mainly from partnerships. In early 2024, the company signed deals with Eli Lilly and Novartis: Lilly paid 45 million dollars upfront with up to 1.7 billion dollars (roughly 1.5 billion euros) tied to milestones, while Novartis paid 37.5 million upfront with up to 1.2 billion dollars (roughly 1.07 billion euros) in potential milestone payments for an initial three targets, a collaboration expanded in 2025. Johnson & Johnson is also a partner. The big sums only flow once the AI-designed compounds make progress in the clinic, and that is exactly where Isomorphic is not yet.

According to Bloomberg, the company has yet to test a drug candidate of its own in humans. In early 2025, Hassabis announced that AI-designed drugs would be in clinical trials by the end of 2025. In January in Davos, he pushed the target to the end of 2026. Whether it holds this time will be decided in the coming weeks. For investors, this means they are valuing a platform, not a product. In a biotech industry where roughly 90 percent of all compounds entering clinical trials fail, a 50 billion dollar valuation before the first Phase 1 study is unprecedented. For comparison: Recursion Pharmaceuticals, the publicly listed pioneer of AI drug discovery with several candidates in the clinic, is worth a fraction of that on the Nasdaq.

“One Thousandth of the Problems”

The skepticism comes, of all places, from a partner. Eli Lilly CEO Dave Ricks said in the spring on Derek Thompson’s podcast that language models are “not particularly good” at biology and chemistry because “they’re trained on the human language, not on the language of chemistry, physics, and biology.” A machine can “predict things pretty well, like predicting the structure of a protein,” Ricks said according to Fortune, “but that is one maybe 1,000th of the kind of problems we face in drug discovery.” In the language of biology, humanity is “sort of like a toddler.” Ricks therefore favors many narrow models for individual prediction problems over one engine that takes over the entire process. That Lilly still partners with Isomorphic shows the industry’s stance: buy in, without buying the grand narrative.

In Isomorphic’s favor: with AlphaFold 3 and IsoDDE, it has models that predict not only protein structures but also their interactions with drug molecules, and the pharma industry has committed to exactly this technology with billion-dollar contracts. Alphabet, for its part, holds a stake in Isomorphic that, should the round close at 50 billion dollars, would be worth more on paper than most acquisitions in the company’s history. Whether it turns into medicines will only become clear in the clinic. The first patient is still waiting.