AI Drug Discovery: Who’s Actually Delivering Clinical Assets?

AI drug discovery is finally starting to produce actual clinical assets, not just hype. For VCs and public equity analysts, the job is to figure out which platforms can turn computing power into a durable pipeline. This report breaks down the main players, looking at their clinical programs and pharma alliances to separate the pure platform stories from the biotechs with real assets.

Mapping Clinical-Stage AI Drug Discovery Platforms

The only real test of an AI drug discovery platform is whether it can get a drug candidate into human trials. Plenty of firms say they use AI for target ID or lead optimization, but only a handful have actually gotten a novel molecule past the preclinical stage and into Phase I. Getting into the clinic is the first major de-risking event, offering a real line of sight to commercialization and proving the platform’s predictions work in actual biology, not just a simulation.

Recursion Pharmaceuticals: Phenomics-Driven Pipeline

Recursion’s whole game is its phenomics engine, which uses AI to make sense of huge biological datasets from its own automated wet labs. Their strategy is to run massive experiments to find new biological connections and drug candidates for a wide range of diseases. Their strong partnerships are key to their funding and market position. For example, Recursion’s work with Nvidia gives them the raw compute power they need to train their models and process data faster, a practical necessity for this kind of work. While Nvidia put $50 million into Recursion in July 2023, they sold off that equity by December 2025, though the two continue to work together on the tech side. Recursion Pharmaceuticals Nvidia partnership details The pipeline itself has seen both progress and setbacks. Their AI-discovered molecule REC-994 for cerebral cavernous malformation was stopped in May 2025 because the SYCAMORE trial’s long-term data didn’t show positive trends or meaningful improvement on MRIs or functional outcomes. ClinicalTrials.gov REC-994 status Recursion’s core advantage is that it generates its own proprietary biological data to feed its AI, creating a data moat that’s very hard for anyone else to build.

Insilico Medicine: Generative AI from Target to Clinic

Insilico was an early mover in using generative AI for both finding new drug targets and creating molecules for them, all with the goal of radically shortening the discovery timeline. Their Pharma.AI platform is built to do just that: first identify a target, then design a novel small molecule to hit it. This “AI-native” process is meant to cut out the dead ends and slow steps of old-school R&D. They’ve made real headway. Insilico now has several AI-generated assets in the clinic, but the main story is their lead program, Rentosertib (formerly INS018_055). It’s a small molecule inhibitor for idiopathic pulmonary fibrosis (IPF) that was discovered and designed by their AI, and it’s now in Phase III trials. Insilico Medicine INS018_055 Phase II announcement Getting a completely AI-native drug into late-stage development is a powerful piece of evidence that their platform works as advertised, showing AI can truly reshape how drugs are invented.

Exscientia: Precision Drug Design and Strategic Alliances

Exscientia uses AI to design small molecules with better properties, focusing on precision to increase a drug’s effectiveness and reduce its toxicity. It’s a tight loop of AI-driven design followed by immediate experimental lab work to create better drug candidates, faster. Their business model leans heavily on co-development deals with big pharma companies, which lets them share the enormous risks and rewards of development. This gives them funding and access to the deep clinical and regulatory experience of their partners. For instance, their AI-designed serotonin 5-HT2A receptor inverse agonist for neuropsychiatric disorders, DSP-0038, is in Phase I through a partnership with Sumitomo Pharma. Exscientia pipeline updates But they also have their own assets, like EXS74539, a wholly owned LSD1 inhibitor for cancer that’s in Phase I. The fact that Exscientia can land these kinds of partnerships is a strong signal that major players believe in their AI platform.

Distinguishing Platform Technology Plays from Asset-Rich Biotechs

So, as an investor, what are you actually buying? The AI platform itself, or the drugs it creates? It’s the most important question to ask. Companies like Recursion, Insilico, and Exscientia are hybrids: the platform is the key differentiator, but the clinical pipeline is what provides the proof and the potential payout.

• Platform Technology Plays: These companies make money by licensing their AI or striking co-development deals. The value is in the platform’s claimed efficiency and ability to find novel drugs. For them to succeed, the tech has to be scalable and consistently produce good candidates.
• Asset-Rich Biotechs: These companies might use AI, but at the end of the day, they’re valued on the strength of their drug pipeline. The AI is just a more efficient tool for building a portfolio. The valuation rests on clinical data. As more AI-discovered molecules get into Phase II, the market is clearly maturing. This flood of new data gives investors something concrete to work with, allowing them to use actual clinical evidence, not just a story, to predict commercial viability and potential exit multiples. The companies that are successfully raising money are the ones who can show assets moving through the clinic, with published outcomes and a clear path to getting reimbursement from payers.
  

  Methodology and Source Note

  All the information in this report comes from cross-referencing public data. We rely on the FDA’s ClinicalTrials.gov database, quarterly financial filings from the companies themselves, and their official corporate pipeline charts. Using this method, we can verify all clinical phase data and partnership details are current. We watch these sources constantly to keep our analysis of the AI drug discovery venture space up to date. It’s a fast-moving field. FDA ClinicalTrials.gov official website