Formation Bio raises $600M drug development fund to solve the clinical trial bottleneck AI is about to make critical

Formation Bio's $600M drug development fund

Ben Liu's core argument is that drug discovery is no longer the bottleneck in pharma — drug development is. The number of discovered drug candidates has doubled over the past decade, even before AI. Approved drugs have stayed flat at roughly 50 per year. The constraint is the cost and time of proving a drug works in humans.

Formation Bio doesn't do discovery internally. It acquires clinical-stage drug candidates and provides both the capital and the operational infrastructure to run trials. A Phase 2 trial runs $60M to $80M or more. A full Alzheimer's program can cost $500M per shot. With AI compressing discovery costs to a few million dollars and falling, Liu argues that gap is about to become acute.

“Drug discovery is getting more and more efficient, and the fundamental bottleneck of bringing new medicines to patients is the cost and time of drug development... A clinical trial can cost a few $100,000,000. Like a phase two clinical trial can be 60 to 80 million... If we don't develop more capabilities around drug development, we're not going to translate all these discoveries into new medicines.”

The fund structure

The $600M fund is asset-level, not company-level. Formation Bio takes individual drug assets into a NewCo, funds the trial, and takes upside through royalties, milestones, and equity. The logic is that most biotechs with a lead asset are forced to neglect their second, third, and fourth candidates for lack of capital. Formation offers to take those shots.

The valuations are asymmetric. A discovered drug is worth little until clinical proof of concept. Post-Phase 2, the heuristic Liu uses is 1x peak annual sales. Post-Phase 3, it's 2-3x. That's the window Formation is targeting.

Why biotechs sell before becoming Roche

Liu frames the acquire-before-scale dynamic as a math problem. If a Phase 2 has a 30% success rate and costs $60M, and a pharma company offers $5B on success, the rational move is to sell rather than run the riskier Phase 3. A founder with three shots should sell the winner. Only once you can take more than ten shots does it make sense to hold and compound. The fund is designed to give Formation — and its biotech partners — enough shots to run that math differently.

AI as a force multiplier on throughput

A traditional pharma company runs 100,000 people because each drug asset requires a dedicated team. Liu says Formation is building AI systems that can exhaustively evaluate every drug-target-indication pair, design trials, and manage development across 10, 15, or 50 assets simultaneously with a team of a few hundred people. The R&D head is a former Biogen executive who developed the first genetic medicine for ALS and the first disease-modifying Alzheimer's drug, paired with AI systems handling the coverage that would otherwise require duplicated headcount.

The commercial picture Liu sees ahead is straightforward: AI compresses discovery costs further, candidate volume rises, and the clinical development bottleneck gets worse unless capacity expands. Formation's bet is that the fund combined with its AI infrastructure can become that capacity — and compound into a drug pipeline large enough to sustain a durable pharma company, something the industry hasn't produced since Amgen and Vertex in the 1980s.