Simon Kohl, a key figure behind DeepMind’s Nobel Prize-winning AlphaFold project, is taking AI-driven drug discovery in a bold new direction. His new startup, Latent Labs, has emerged from stealth with $50 million in total funding to build generative AI models that create entirely new proteins—designed for specific therapeutic needs rather than constrained by nature’s existing repertoire. Unlike many AI drug discovery startups, Latent Labs won’t develop its own drugs but will act as a service provider to pharmaceutical companies, offering a customizable AI-driven platform for protein design.
Key Points:
- The company offers a platform for pharma company to generate custom proteins rather than developing drugs in-house.
- Backers include Radical Ventures, Sofinnova Partners, Google’s Jeff Dean, and Cohere’s Aidan Gomez.
- Latent Labs aims to make biology ‘programmable,’ potentially cutting drug development timelines.
Kohl believes the next frontier in AI-driven drug development isn’t just predicting protein structures but designing new ones from scratch. “We think we’re at a point where generative modeling allows us to go beyond what nature has given us,” Kohl said. “In a perfect world, drugmakers could input their disease target, and AI would generate optimized candidates at the push of a button.”
Unlike other AI-driven biotech startups, Latent Labs isn’t building its own drug pipeline. Instead, it is positioning itself as a service provider, offering pharmaceutical companies AI tools to create proteins with precise therapeutic properties. This approach contrasts with DeepMind’s own drug discovery spinout, Isomorphic Labs, which is focused on small-molecule drugs.
The startup’s $50 million in funding includes a $40 million Series A round co-led by Radical Ventures and Sofinnova Partners, with additional backing from AI heavyweights like Google’s chief scientist Jeff Dean, Cohere co-founder Aidan Gomez, and ElevenLabs founder Mati Staniszewski. Investors see Latent Labs’ “de novo” protein design capabilities as a transformative shift for computational biology. “Such a capability has never before been possible, one which can benefit humanity in a profound way,” said Radical Ventures partner Aaron Rosenberg.
Latent Labs enters a competitive but rapidly expanding market, alongside companies like Generate:Biomedicines and Cradle Bio, which also use AI for protein engineering. However, while many firms focus on optimizing existing proteins, Latent Labs is betting on its ability to generate entirely new ones. If successful, this could unlock new classes of medicines that were previously out of reach, potentially reducing drug development timelines and costs.
With a growing team of AI researchers from DeepMind, Google, and Microsoft, Latent Labs is now focused on scaling its technology and forming partnerships with pharmaceutical companies. The company’s success will depend on how effectively it can integrate its AI-generated proteins into real-world drug development—but if it delivers, it could change how medicines are designed from the ground up.