In November 2024, Sir Demis Hassabis walked onto a stage in Stockholm to accept the Nobel Prize in Chemistry. The award honored AlphaFold 2, the protein-structure model that his team at Google DeepMind had built and that, in the four years since its release, had been used by more than two million scientists in 190 countries to predict the shape of nearly every protein known to biology.
Eighteen months later, Hassabis has converted that Nobel-winning science into a balance sheet.
On Tuesday, May 12, the Alphabet-owned spinout Isomorphic Labs announced a $2.1 billion Series B funding round led by Thrive Capital. New investors include Abu Dhabi's MGX, Singapore's sovereign wealth fund Temasek, and the United Kingdom Sovereign AI Fund. Existing backers Alphabet, GV, and CapitalG also participated. The round, Bloomberg's Mark Bergen and Rebecca Torrence reported, is the second time the four-year-old company has taken outside capital. It is also one of the largest private financings ever recorded in AI drug discovery.
The capital pushes Isomorphic's total funding base past $2.6 billion, according to industry estimates. The company's first drug candidates, designed end-to-end inside its proprietary AI engine, are expected to enter human trials by the end of this year.
The Engine the Money Will Build
Isomorphic's signal product is called the Isomorphic Drug Design Engine, or IsoDDE. The company announced in February 2026 that IsoDDE doubled the performance of AlphaFold 3 on a protein-ligand structure prediction generalization benchmark. Where physics-based methods take weeks of compute time and a team of medicinal chemists to evaluate a single binding affinity, IsoDDE predicts the same value with higher accuracy at a small fraction of the cost.
More important for drug discovery, IsoDDE identifies new binding pockets on target proteins using only the amino acid sequence. That capability matters because the limiting factor in modern drug discovery is not synthesizing molecules; it is finding the right molecule to synthesize. AlphaFold 2 solved the protein-folding problem. IsoDDE is the first system from the AlphaFold lineage to solve the drug-design problem at industrial scale.
The Series B proceeds will fund three concrete priorities, according to Isomorphic's announcement: scaling the engine, expanding the pipeline of therapeutic programs toward the clinic, and hiring across AI research, computational chemistry, clinical operations, and engineering. The company has consistently described its strategy as building infrastructure first, then drug candidates, then a fully AI-native pharmaceutical company.
The Investors Tell You What This Is About
The composition of the round is worth studying. Two-thirds of the named outside investors are sovereign wealth vehicles.
| Investor | Origin | Investment Style |
|---|---|---|
| Thrive Capital (lead) | New York | Growth-stage tech |
| Alphabet | United States | Parent company |
| MGX | Abu Dhabi | Sovereign AI fund |
| Temasek | Singapore | Sovereign wealth fund |
| UK Sovereign AI Fund | United Kingdom | Sovereign AI vehicle |
| GV | United States | Alphabet venture arm |
| CapitalG | United States | Alphabet growth fund |
Sovereign wealth funds do not typically lead drug discovery rounds. They lead infrastructure rounds. Their presence on the Isomorphic cap table reflects a thesis that AI-designed drugs are no longer a research bet; they are a national-strategy bet. The Gulf, Singapore, and the United Kingdom are each, in different ways, racing to position themselves as homes for AI-native pharmaceutical industries. Putting capital into the company most likely to ship the first AI-designed drug is a hedged way to enter that market without picking a host country.
For Alphabet specifically, the round has a different significance. Alphabet has been Isomorphic's primary backer since the company spun out of DeepMind in 2021. Bringing in Thrive Capital as the lead, with sovereign co-investors, dilutes Alphabet's effective ownership in exchange for a much larger capital pool and validation that Isomorphic is no longer a DeepMind side project.
The Pharma Partners Are Already Watching
Isomorphic does not yet sell drugs. It sells a service: research partnerships with major pharmaceutical companies, in which Isomorphic's AI engine designs candidate molecules and the pharma partner takes them through clinical development. The partnership roster is the closest thing the AI drug discovery industry has to a leaderboard.
- Novartis: Multi-target strategic collaboration announced in 2024
- Eli Lilly: Multi-target deal for cardiometabolic and oncology programs
- Johnson & Johnson: Multiple targets across therapeutic areas
Each of those agreements carries upfront fees, milestone payments, and royalties on any drugs that reach the market. Isomorphic has not disclosed the total contracted value of its pharma partnerships, but the company's last reported deal terms valued the Lilly and Novartis partnerships at hundreds of millions in upfront and milestone payments alone. Industry estimates put Isomorphic's pre-revenue contracted value above $2 billion across its three named partners.
The Series B announcement coincided with a stated goal that Isomorphic's first internal drug candidates, not just partner candidates, will reach human trials by the end of 2026. That timeline, if hit, will make Isomorphic the first AI-native company to take an AI-designed drug into Phase 1 from its own pipeline.
The Other Side
Not everyone is convinced that AI drug discovery has earned this kind of capital.
The skeptical view, articulated most forcefully by veteran biotech analysts and the editorial team at Endpoints News, runs as follows. AI drug discovery has been the recipient of more capital in 2024 and 2026 than any new biotech subsector since immuno-oncology, and yet the field has produced exactly zero approved drugs.
Three cautionary tales are usually offered. Recursion Pharmaceuticals, the most public AI-native drug discovery company, has spent more than a decade and over $1 billion to advance a small pipeline that has yet to deliver a commercial drug. BenevolentAI, once worth roughly two billion dollars in market value, lost three-quarters of its market capitalization between 2022 and 2024. Exscientia, which entered Phase 1 with its first AI-designed candidate in 2020, terminated the program before Phase 2 readouts and was acquired by Recursion in 2024 at a steep discount.
The skeptical case against Isomorphic specifically is that AlphaFold's success in protein structure prediction does not necessarily transfer to the harder problem of drug efficacy. Predicting a protein's shape is a problem with a known answer. Predicting whether a small molecule will safely cure a disease in a human body is not. The former is computer science; the latter is biology.
A more technical critique came from medicinal chemists who reviewed Isomorphic's February IsoDDE paper. While the doubled performance on the protein-ligand benchmark is real, the benchmark itself measures generalization on known target classes. The grand challenge of drug discovery is novel targets where no validated binders exist. Whether IsoDDE crosses that gap remains, by Isomorphic's own admission, an open question.
There is also a regulatory dimension. The FDA has approved exactly zero drugs whose design rationale was solely an AI prediction. Every approval to date has required wet-lab confirmation at every stage. If Isomorphic's first internal candidate reaches Phase 1 in 2026, it will move through the same approval gauntlet as a traditionally designed drug, with the same approximately one-in-ten odds of reaching the market.
The Bottom Line
Isomorphic Labs is the bet that wins the most when AI works in drug discovery and loses the most when it does not. The Series B prices that bet at $2.6 billion in committed capital, with backers ranging from Alphabet's internal funds to the largest sovereign wealth pools in the Gulf, Asia, and Europe. The asset on the other side of that capital is a single AI engine whose first internal candidates are months away from a human body.
What the round really buys is time. Drug discovery is a 10-to-15-year cycle. Isomorphic now has enough cash to run its pipeline for a decade without raising again, and to do so in an environment where every major pharma company will be choosing AI-native partners over the next five years. If IsoDDE works, the company that owns it will be the most valuable asset in pharmaceutical research. If it does not, Isomorphic will be the most expensive science experiment ever funded, and the lesson the industry takes from the failure will reshape the field for the next decade.
As Hassabis himself framed it in the funding announcement: the goal is to bring much-needed treatments to millions of patients globally. The next 18 months will determine whether that ambition is realized in Stockholm again, or only in another funding deck.
Sources
- DeepMind Spinout Isomorphic Labs Raises 2.1 Billion to Design Drugs With AI (Bloomberg, May 12, 2026)
- Isomorphic Labs secures 2.1 Billion funding to scale its AI drug design engine (PR Newswire / Isomorphic Labs, May 12, 2026)
- Isomorphic Labs announces Series B investment round (Official Announcement) (Isomorphic Labs, May 12, 2026)
- Five burning Qs for Isomorphic Labs after 2.1B raise for AI biotech (Endpoints News, May 12, 2026)
- Isomorphic Labs raises 2.1 billion Series B for AI drug discovery (Quartz, May 12, 2026)
- Isomorphic Labs Announces 2.1B Funding to Scale its AI Drug Design Engine (The AI Insider, May 13, 2026)
- Sir Demis Hassabis, PhD (Isomorphic Labs Profile) (Isomorphic Labs, retrieved May 13, 2026)
- Demis Hassabis in 2026: How DeepMind Is Turning AlphaFold Into a Drug Discovery Engine (StartupHub.ai, May 11, 2026)
- Isomorphic Labs (Wikipedia, retrieved May 13, 2026) (Wikipedia, May 2026)