Demis Hassabis, CEO of Isomorphic and Google DeepMind
© 2025 Bloomberg Finance LP
Isomorphic, which formed as an Alphabet company in 2021, has a massive cash horde, but it’s tight-lipped on what drugs it might bring to the clinic.
On Tuesday, Isomorphic Labs, the Alphabet-founded firm that uses AI to accelerate drug development, said that it had raised $2.1 billion from investors led by Thrive Capital. That’s a boatload of cash for a drug developer, and represents the second largest ever fundraise for a biotech company behind only Altos Labs, according to trade publication Endpoints News. And it’s the latest sign of the market’s high hopes that AI can fix the costly, lengthy drug development process.
“We’re aiming to redefine the way we create new medicines,” Isomorphic President Max Jaderberg tells Forbes, adding that the new funding represents “a lot of validation of what we’ve been building out the past four-and-a-half, almost five, years.” The investment, which is at an undisclosed valuation, follows Isomorphic’s first outside funding of $600 million last year.
London-based Isomorphic is best known for AI model AlphaFold, which predicts protein structures, for which CEO Demis Hassabis (who also remains CEO of Google DeepMind) won the 2024 Nobel Prize in Chemistry. The model’s latest version, AlphaFold 3, came out in May 2024, and works with the array of building blocks of drugs, including small molecules, peptides and antibodies, as well as proteins. The company has since built off this work by developing what it calls Isomorphic Labs Drug Design Engine, or IsoDDE. Jaderberg calls IsoDDE “like half a dozen AlphaFold breakthroughs,” with the ability to, among other things, determine how well a therapy binds to its target and whether there may be toxic side effects.
“The really cool thing about the Iso drug design engine is that it is not specialized to a disease area. It’s largely agnostic in that way, which speaks to our big mission of solving all disease,” says Jaderberg, who first joined Google back in 2014 and was previously Isomorphic’s chief AI officer.
The bigger question than the technology, though, is what Isomorphic will be able to do with it. And on that Jaderberg is cagey.
The firm has signed partnerships with big pharma companies Novartis and Eli Lilly that could be worth up to $3 billion combined. More intriguingly, it is also designing therapeutics in house—a big move for a company rooted in Big Tech.
Jaderberg says that Isomorphic has been expanding its pipeline of therapies due to its success with the drug design engine. Its internal work is now focused on oncology, immunology and inflammation. “The programs we are going after are not fast-follower programs,” Jaderberg says. “We are thinking about going after these big zero-to-one problems where maybe the rest of the world has struggled to produce good medicines for patients or we can change the standard of care.”
Jaderberg says that Isomorphic intends to take those drug candidates to clinical trials itself, though he declined to discuss the company’s lead therapies, or what the timeline for getting into the clinic would be. (At an event at this year’s World Economic Forum in January, Hassabis gave a timeline of the end of 2026, a full year delay from what he previously said.)
Whether Isomorphic will ultimately sell off those assets, license them to pharmaceutical companies or choose to become a more traditional drug company by marketing the therapeutics it creates also remains an open question. Jaderberg says that he considers each drug as essentially “an individual business,” so the answer could be different for each one.
The field has changed a lot since Recursion launched back in 2013, and there are many more AI drug developers now. Among the newer crop: Chai Discovery, which is valued at $1.3 billion and has a deal with Lilly, and Manifold Bio, which has raised $40 million and signed one with Roche.
“It’s so exciting to see how this AI for drug discovery space has blossomed since we started,” Jaderberg says. “When we started the company, changing the way we do drug design with AI was a hypothesis. … It’s moved from a hypothesis to this is real and we know this works.”
Now all Isomorphic needs to do is show that a drug it designed with AI can actually get to patients.
MORE FROM FORBES