Demis Hassabis's Isomorphic Labs: $2.7B Raised, Trials Due This Year

Isomorphic Labs, the AI drug-discovery company created by Google DeepMind, raised $2.1 billion in a Series B round in May 2026, according to Bloomberg, pushing its total outside capital to $2.7 billion across two rounds in just over 14 months. The company now holds nearly $3 billion in committed pharmaceutical milestones from Eli Lilly and Novartis, a new design engine that more than doubled the accuracy of AlphaFold 3, and a first-in-human cancer trial on the calendar for 2026.

A $2.7 billion capital stack

Isomorphic Labs received its first external capital in March 2025: a $600 million Series A led by Thrive Capital, with Alphabet and its venture arm GV participating. That round carried no disclosed valuation; Alphabet has always held a majority stake since the spinout was incorporated in 2021.

Fourteen months later, on May 12, 2026, Isomorphic closed its Series B at $2.1 billion. Thrive Capital returned as lead, joined by Alphabet, Abu Dhabi's MGX, Singapore's Temasek, and the UK's Sovereign AI fund, Bloomberg reported. No post-money valuation was disclosed. The participation of sovereign wealth funds from three continents reflects a pattern of governments co-investing in AI infrastructure they view as having national-security implications beyond pure financial returns.

In a Fortune interview published around the close of the Series B, Hassabis described the company's ambition in deliberately expansive terms: "It's actually about building the platform to solve the overall problem, solving all disease, right. We don't measure ourselves by one individual program." He framed Isomorphic's mandate as building general-purpose biology infrastructure rather than optimising for any single drug candidate, a posture the $2.1 billion round is designed to fund.

Two pharma deals and a new design engine

Isomorphic's commercial foundation predates both funding rounds. In January 2024, the company announced simultaneous drug-discovery collaborations with Eli Lilly and Novartis. The Lilly deal included a $45 million upfront payment and up to $1.7 billion in performance-based milestones; the Novartis deal provided $37.5 million upfront and up to $1.2 billion in milestones. In February 2025, Novartis expanded the collaboration to add three additional research programs. Combined, the two partnerships carry nearly $3 billion in milestone potential, excluding royalties on any eventual drug sales.

The deal economics matter because they establish what large pharmaceutical firms are willing to pay for AI-directed early-stage discovery, a phase that traditionally costs hundreds of millions of dollars and takes three to five years before a single compound advances to a clinical trial. Isomorphic's proposition is to compress that timeline and improve hit rates by computationally designing molecules rather than screening chemical libraries.

The technical credibility behind that proposition advanced significantly in February 2026, when Isomorphic published a 27-page technical report on IsoDDE, its Isomorphic Drug Design Engine. On the hardest protein-ligand structure-prediction benchmarks, IsoDDE achieved 50% accuracy, compared with 23.3% for AlphaFold 3, the system that contributed to Hassabis's Nobel Prize. Columbia University computational biologist Mohammed AlQuraishi characterised the advance as comparable in magnitude to an entirely new AlphaFold version, in assessments that circulated widely in the structural-biology research community following the report's release.

The clinical trial reckoning

Fundraising figures and benchmark scores are proxies. The real test is whether an AI-designed molecule survives first contact with a human body. Isomorphic currently has 17 drug programs underway across oncology and immunology, with plans to scale to hundreds, according to company president Colin Murdoch. In July 2025, Murdoch said first-in-human trials were "very close."

At the World Economic Forum in Davos in January 2026, Hassabis confirmed that an AI-designed cancer drug candidate is scheduled to enter Phase 1 clinical trials before the end of 2026, a revision from an earlier target of end-2025. Phase 1 trials test safety and dosing in a small cohort of patients rather than efficacy; the result that matters commercially is whether the compound advances to Phase 2 and, ultimately, proves more effective than what conventional pipelines would have found through standard screening methods.

The $2.1 billion Series B is sized, in part, to carry the company through that multi-year trial arc. Drug development timelines are measured in years rather than product cycles, and the sovereign-fund investors who joined the round appear comfortable with that horizon. Whether AI-directed design actually improves Phase 1-to-2 progression rates is the data point the entire pharmaceutical industry is waiting on.

What it means

Isomorphic Labs has built, in three years, the most credentialed and best-capitalised pure-play AI drug-discovery company: Nobel Prize-backed science, $2.7 billion raised, nearly $3 billion in pharma commitments, and a design engine that materially outperforms its predecessor. The company's structural position is unusual; Alphabet's majority ownership insulates it from quarterly return pressure while sovereign wealth co-investors validate the national-security framing of AI-led drug discovery. The open question is not whether Hassabis can raise capital or publish benchmarks. It is whether the first AI-designed compound to enter a human trial performs well enough to rewrite the economics of pharmaceutical R&D at scale.

Sources

• Bloomberg: DeepMind Spinout Isomorphic Labs Raises $2.1 Billion to Design Drugs With AI
• Bloomberg: Alphabet's AI Drug Unit Raises $600 Million From OpenAI Backer
• Fortune: Demis Hassabis on his rush to solve all disease and Isomorphic's new $2.1 billion
• Isomorphic Labs: Kicks off 2024 with two pharmaceutical collaborations
• Time: Google DeepMind's AlphaFold 3 Could Transform Drug Discovery

Editorial standards: every claim is sourced. Tips: [email protected]