Google's Isomorphic Labs is reportedly looking for more than $2 billion, a sign that AI drug discovery is moving into a far more expensive phase.
Google's life sciences bet is no longer just a research story sitting beside DeepMind. Isomorphic Labs, the Alphabet-owned company built to turn AI breakthroughs into new medicines, is seeking to raise more than $2 billion in fresh funding, according to Bloomberg, putting one of Big Tech's most ambitious biotech projects into the center of the AI capital race.
The number matters because drug discovery is not a chatbot market. It does not reward fast demos in the same way, and it does not produce enterprise value just because a model performs well in a benchmark. It is slow, regulated, expensive and unforgiving. If Isomorphic can raise at that scale, investors are effectively saying that the next major AI business may not only live in software subscriptions or cloud infrastructure, but in the long road from molecular design to clinical trials.
Isomorphic was launched in 2021 as a DeepMind spinout under Demis Hassabis, the DeepMind co-founder who also leads Google DeepMind. Its foundation came from the same scientific progress behind AlphaFold, the system that helped make protein structure prediction one of the clearest examples of AI producing useful scientific work. That heritage gives Isomorphic something many AI-biotech startups spend years trying to build: credibility with researchers, access to deep technical talent and a parent company with the balance sheet to keep funding work before revenue becomes predictable.
That last point is becoming more important. In March 2025, Isomorphic raised $600 million in its first external funding round, led by Thrive Capital with participation from Alphabet and GV. It also signed major research collaborations with Eli Lilly and Novartis that carried potential milestone payments of nearly $3 billion combined. Those deals gave the company a commercial path, but a larger raise would suggest Alphabet is preparing Isomorphic for a much broader role than partnership work alone.
The AI market has spent the last few years chasing visible products. Chatbots, coding assistants, search tools and enterprise copilots were easy for investors to understand because customers could start using them quickly. AI drug discovery is different. A promising molecule is not a product. A partnership announcement is not proof of approval. A model that predicts interactions well still has to survive laboratory testing, toxicology, human trials and regulatory review.
That makes the funding logic sharper. The companies that win in this market will need more than good models. They will need proprietary biological data, tight feedback loops between software and wet labs, enough compute to keep improving their systems and enough money to keep programs alive through long clinical timelines. That combination naturally favors companies with either very deep-pocketed investors or strategic parents that can tolerate years of uncertainty.
This is where Isomorphic has a structural advantage. Independent AI-biotech startups can be excellent technically, but they are often raising in markets where venture firms want clearer milestones and cleaner comparisons. Google can afford to think differently. Alphabet already spends heavily on AI infrastructure, has world-class research assets through DeepMind and can support a company whose payoff may arrive over a decade rather than a quarter.
The risk is that AI drug discovery becomes a winner-take-most market, not because one model solves biology, but because only a handful of platforms can finance the full stack. If the best data, pharma partnerships, research talent and compute cluster around the same few players, smaller companies will have to prove they own something truly differentiated. A narrow disease focus, a unique dataset or a better clinical development strategy may matter more than broad claims about AI-powered discovery.
Why This Matters Beyond Biotech
For startup founders and investors, the Isomorphic raise points to a broader shift in AI funding. The first wave rewarded companies that could package general-purpose models into tools people could buy immediately. The next wave may reward companies willing to apply AI inside sectors where the economic prize is huge but the path is slower and more technical. Healthcare, energy, materials science and advanced manufacturing all fit that pattern.
That shift will test the patience of the AI investment cycle. It is one thing to value a software company on revenue growth and usage. It is another to value a drug platform on the probability that a candidate designed with AI reaches approval years from now. The upside can be enormous, but the proof will come through clinical progress, not dashboard metrics.
It also changes how Big Tech competes with startups. Google does not need Isomorphic to look like a normal venture-backed company. It can function as a strategic platform, a scientific brand and a possible long-term business line at the same time. For independent startups, that raises the bar. They will need to show why a pharma company should work with them instead of a platform tied to DeepMind's research engine and Alphabet's capital base.
None of this means AI is close to replacing the messy work of drug development. Biology has a way of humiliating elegant software. But the funding signal is clear. Investors are beginning to underwrite AI companies that operate in harder markets, where the prize is not faster content or cheaper coding, but a better shot at creating medicines that would otherwise take years and billions to develop.
The next thing to watch is not only whether Isomorphic closes the round, but what it does with the money. Hiring more scientists is one thing. Advancing internal drug programs into the clinic is another. If Isomorphic can move from celebrated research platform to credible biotech engine, the AI boom will have entered a more serious and much more expensive chapter.
Also read: DeepSeek's funding talks show China's AI race is getting expensive • Three Mile Island is becoming an AI power bet • Google DeepMind has raised the bar for AI math reasoning
