Google's Secret Weapon Just Raised More Than Most Biotechs Are Worth

Think of it this way: traditional drug discovery is like trying every key in a hardware store to see which one opens a particular lock. Isomorphic's AI looks at the lock and designs the key from scratch.

The Pharma Partnerships That Matter

A fancy AI platform is one thing. Getting Eli Lilly and Novartis to bet billions on it is another.

In January 2024, Isomorphic inked deals with both pharma giants worth a combined $3 billion in potential milestone payments. Lilly paid $45 million upfront with up to $1.7 billion in milestones. Novartis put down $37.5 million with up to $1.2 billion on the table. Both deals focus on small molecule drugs aimed at targets that have historically been considered "undruggable," the biological equivalent of picking locks that no one has been able to crack.

The early results seem to be working. Novartis expanded its collaboration in February 2025, effectively doubling the number of programs from three to six. By early 2026, both partnerships had reportedly moved beyond initial discovery into preclinical candidate stage, with several programs entering the final stretch of testing required before human trials can begin.

Isomorphic is now staffing up its clinical development team, hiring regulatory experts and trial designers. The company has said it plans to put its first AI-designed molecules into Phase 1 human trials by late 2026. If that happens, it would be a landmark moment: proof that a drug designed primarily by artificial intelligence can make the leap from silicon to skin.

Why $2.1 Billion, and Why Now?

The timing isn't accidental. AI drug discovery funding has been on a tear. But this round stands apart, both in size and in what it signals.

First, the investor mix tells a story. Thrive Capital coming back to lead (for the second consecutive round) signals deep conviction, not tourist money. The sovereign wealth funds from Abu Dhabi, Singapore, and the UK suggest that governments now view AI-driven drug development as critical infrastructure, similar to how they invest in semiconductors or energy.

Second, the capital gives Isomorphic a structural advantage. Training massive AI models costs a fortune. Generating the biological data to feed those models costs even more. With $2.1 billion in fresh cash on top of last year's $600 million, Isomorphic can outspend virtually every competitor in the space on compute, data generation, and talent. In AI, scale tends to compound; more data makes better models, which attract better partners, which generate more data.

Third, analysts note that the real test is still ahead. No AI-discovered drug has ever received FDA approval. Several have entered clinical trials; some have already failed. Isomorphic's Drug Design Engine may be impressive on benchmarks, but biology has a nasty habit of humbling even the best predictions. The gap between "works on a computer" and "works in a human" remains the great unsolved problem of the field.

The Bigger Picture

Isomorphic Labs now sits at a fascinating inflection point. It has the pedigree (Nobel Prize-winning science), the partnerships ($3 billion in pharma deals), the capital ($2.7 billion raised to date), and a clinical timeline that could produce real human data within months.

But it also has no approved drugs, no clinical results, and a valuation that remains undisclosed, probably because the number would make traditional biotech investors spit out their coffee.

Hassabis has talked publicly about compressing early drug discovery from years down to months. If Isomorphic delivers on even a fraction of that promise, it could reshape how the entire pharmaceutical industry operates. If it doesn't, this will go down as the most expensive science experiment in biotech history.

Either way, the bet has been placed. And at $2.1 billion, it's the biggest one AI drug discovery has ever seen.