The big drugmakers are already in

Generative AI in pharma is not a future trend; it is a set of deals already signed. The largest pharmaceutical companies have spent the last few years building internal AI teams and, more visibly, writing checks to the small firms that specialize in the technology. The “sub-$10 AI drug stock” promo from Alexander Green’s Oxford Club points at one of those small firms, Absci, and the deal sheet behind it is a clean map of how this market actually works.

The pattern is consistent: a small company develops a generative platform, a big drugmaker pays for access, and the deal is structured around milestones and royalties. That structure, not the technology itself, is what an investor in this space is really buying, and it is worth understanding before any of the tickers.

How the deals are structured

Absci’s roster shows the template. Merck signed a drug discovery deal in 2022 worth up to about $610 million in milestone payments. AstraZeneca followed in 2023 with a deal worth up to about $247 million to apply the platform to cancer drugs. Eli Lilly made a $40 million strategic investment tied to the lead program, and AMD put in $20 million alongside a hardware collaboration.

The “up to” is the operative phrase. These deals pay out as the drug program advances, so the headline number is a ceiling, not a check that clears on signing. The pharma partner bears most of the downstream cost of testing and commercialization, while the platform company collects milestones and, eventually, royalties on anything that reaches market. That is the royalty model the promo describes, and it is how this corner of the industry has actually been structured for years. We explain the royalty mechanics in our best royalty stocks piece.

Why pharma buys instead of builds

Big pharma has the money to build its own generative AI, and it is doing so, but it still buys from small specialists. The reason is that the technology moves fast and the talent is scarce. A large drugmaker can sign a deal and get access to a working platform, a trained model, and the team behind it, far faster than it can assemble that capability in-house. The specialist, in turn, gets revenue, validation, and, importantly, access to the partner’s data and biology, which makes its own model better.

That mutual dependence is why the partnership model has taken over this corner of the industry. It is not a sign that the small firms are weak; it is the economic shape of a technology that is still maturing. Our generative AI drug discovery explainer covers the technology underneath the deals.

The timing problem, again

The caveat that follows every generative AI in pharma discussion is timing. A partnership can be signed today and produce its first clinical candidate years from now, and a royalty check arrives only when a drug is on the market. Absci has a lead asset in trials, ABS-201 for pattern hair loss, with proof-of-concept data due in the second half of 2026, but even a successful readout is years from a product on shelves.

That is not a reason to dismiss the technology; it is a reason to price it honestly. The companies worth watching are the ones with candidates in trials, because a deal alone is a promise and a trial is progress. Our antibody drug discovery guide explains why antibodies are the leading edge of this work.

What a win looks like for pharma

From the drugmaker’s side, the payoff is not a stock price but a pipeline. A pharma company that can generate more promising candidates per dollar of research spend gets more shots on goal, and more shots on goal is the entire game in an industry where most programs fail. That is why the internal AI teams exist alongside the external partnerships; they are two routes to the same outcome.

The practical result is that the small specialist firms are not really competing against pharma. They are selling pharma a faster way to fill its own pipeline, and their value is measured in the time and failure they remove from the front of the process. That is the honest frame for the “royalty business” claim the promo makes: the value is real, but it accrues slowly and only after the drugs themselves succeed.

The takeaway

Generative AI in pharma is real, funded, and already embedded in the deal flow of the biggest drugmakers in the world. The partnerships are structured the way they are for a reason, and the reason is that drug development is long and risky and the risk is being shared. For an investor, the useful lens is simple: look past the technology to the deal structure, then past the deal structure to the clinical pipeline, because that is where the returns will actually be decided.

Ready to see the research? Click here to access Alexander Green’s report.

NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.