Why antibodies are the sweet spot for AI

Antibodies are among the most valuable drugs ever made, and they are also the place where AI drug discovery is making the fastest progress. The reason is structural: antibodies are proteins, and proteins are exactly the kind of molecule that AI models can represent digitally and search. Small-molecule drugs are simpler but their chemistry is harder to model; antibodies are complex but their design space is more legible to a machine that has been trained on protein data.

The “sub-$10 AI drug stock” promo from Alexander Green’s Oxford Club points at a company built on this idea. Absci, the firm behind the tease, designs antibodies with a generative AI platform, and its lead asset, ABS-201, is itself an antibody. So the promo is not just about AI drug discovery in general; it is about the antibody corner of it, and that corner has specific mechanics worth understanding.

What an antibody actually is

An antibody is a protein the immune system uses to recognize and neutralize threats, and drugmakers have learned to design them to hit specific disease targets. The appeal is precision: an antibody can be engineered to bind to a particular protein and leave everything else alone, which is why antibody drugs now treat everything from cancer to autoimmune disease.

The catch is that designing one is hard. The space of possible antibody sequences is astronomically large, and finding the one that binds tightly, stays stable in the body, and does not trigger a harmful immune response is a needle-in-a-haystack problem. That is exactly the kind of search a generative AI model is good at, and it is why the antibody corner has become the showcase for AI drug discovery. Our AI powered drug discovery explainer covers the mechanism behind the search.

Engineering the half-life

Absci’s lead asset illustrates a specific design goal: extending half-life. ABS-201 is an antibody engineered to last longer in the body, which matters because a longer-lasting drug needs to be dosed less often and can be more convenient for patients. It targets pattern hair loss, a condition affecting an enormous number of people, and is also being tested in endometriosis.

The hair loss angle is what makes the asset commercially interesting. A well-tolerated antibody for a huge cosmetic-plus-medical market would be a meaningful product, and the June safety readout was encouraging enough to move the stock. The proof-of-concept data is the next test, expected in the second half of 2026. For how the big pharmaceutical companies are adopting this technology, see our generative AI in pharma piece.

The risk specific to antibodies

The antibody corner has its own failure modes, and the biggest is immunogenicity: the risk that a patient’s immune system reacts against the drug itself. An engineered antibody can be perfectly effective in a dish and useless in a person if the body mounts a response to it. That is one reason clinical data matters so much more than platform demos in this space.

There is also competition. In hair loss specifically, Veradermics is already in late Phase 3 with an oral treatment, and an oral pill is a lower bar for patients than an injected antibody. That does not invalidate Absci’s approach, but it does mean the antibody’s advantages, precision and longer half-life, have to be weighed against the convenience of a pill. Our AI drug discovery explainer covers the broader field.

Monoclonal antibodies, the blockbuster class

Most of the antibody drugs that have become household names are monoclonal antibodies, which means they are copies of a single immune cell’s antibody, engineered to hit one specific target. The class has produced some of the best-selling drugs in history, from cancer therapies to autoimmune treatments, and it is the reason antibody discovery attracts so much investment.

That history is also the source of the commercial logic in the Absci pitch. A company that can reliably design a monoclonal antibody against a new target is selling into a proven, high-value market, not a speculative one. The risk is not that the market does not exist; it is the engineering and clinical risk of getting a specific antibody to work in humans, which is exactly the part a platform demo cannot answer.

The takeaway

Antibody drug discovery is the part of AI biotech where the mechanism is clearest and the stakes are most concrete. The companies worth watching are the ones that have moved a designed antibody into human trials, because that is the point where a clever platform becomes a real drug program. Absci is one such company, and its antibody for hair loss is the specific bet the promo is really asking you to make.

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.