Google DeepMind to soon begin human trials of AI-designed drugs
What's the story
Google DeepMind's drug discovery unit, Isomorphic Labs, is gearing up to begin human trials of its AI-designed drugs. The announcement was made by Colin Murdoch, President of Isomorphic Labs and Chief Business Officer at Google DeepMind. He said the company is "getting very close" to starting these trials in an interview with Fortune. This marks a major milestone in the field of pharmaceutical innovation.
Origin story
Isomorphic Labs and its mission
Isomorphic Labs was spun out of DeepMind in 2021, leveraging the power of AlphaFold, a groundbreaking AI system from DeepMind that accurately predicts protein structures. Over time, AlphaFold has evolved to model interactions between proteins and molecules like DNA and drugs. This evolution has greatly enhanced its role in drug discovery, speeding up the process of developing new medicines.
Strategic alliances
Partnerships with Novartis and Eli Lilly
In 2024, Isomorphic Labs partnered with pharmaceutical giants Novartis and Eli Lilly. The company also raised $600 million in its first external funding round led by Thrive Capital in April 2025. These strategic moves are aimed at building a "world-class drug design engine" that integrates AI and pharma expertise for more efficient medicine development.
Aspirations
Company is also working on its drug candidates
Along with supporting existing drug programs, Isomorphic Labs is also working on its own candidates in oncology and immunology for future licensing. Murdoch explained that the company identifies unmet needs and starts its own drug design programs, which they aim to put into human clinical trials. However, he admitted they haven't reached that stage yet but are making good progress toward it.
Success rates
DeepMind's tech could transform pharmaceutical innovation
Currently, pharma companies pour millions into drug development with just a 10% success rate in trials. Murdoch hopes to use AlphaFold's tech to dramatically improve these odds, ideally getting to a point where researchers can be completely confident of a drug's effectiveness before human trials even start. "One day we hope to be able to say — well, here's a disease, and then click a button and out pops the design for a drug," he said.