XtalPi moves its AI-designed hair-loss drug into IND-enabling studies

XtalPi on Friday said its proprietary hair-loss drug program has entered IND-enabling studies, building on two AI-developed scalp-care ingredients already commercialized under its Groland brand.

Most "AI drug" headlines stop at a paper. XtalPi is pushing a hair-loss candidate from cosmetic ingredients into IND-enabling work — the paperwork, safety, and manufacturing grind that decides whether an AI-designed molecule ever reaches the FDA.

On Friday, 2 October 2026, XtalPi said its proprietary hair-loss drug program has entered IND-enabling studies. The PR Newswire page is titled “XtalPi Advances Hair Loss Drug Program into IND-Enabling Studies.” The page stamps Oct. 2, 2026, 18:33 ET, which is 6:33 p.m. Eastern. The dateline is Cambridge, Massachusetts, and Shenzhen, China. The source line is XtalPi Inc. The shares trade as 2228.HK. IND-enabling studies are the lab work a company finishes before it asks the U.S. Food and Drug Administration for permission to test a new drug in people. That request is an Investigational New Drug application, often shortened to IND. The release says pharmacology, safety, and chemistry, manufacturing, and controls work are underway to support a future application. Pharmacology is what the drug does in a living system. Chemistry, manufacturing, and controls, often shortened to CMC, is the proof that each batch can be made the same way, at a known quality. Those lines are the wire's. The page does not say an IND has been filed. It does not print a filing date.

What the candidate is supposed to do. The release describes a topical medicine, one put on the skin, for pattern hair loss. The medical name on the page is androgenetic alopecia. The release says it affects both men and women and can require long-term management. The candidate is described as acting through a lactate–autophagy pathway to reactivate quiescent hair-follicle stem cells. Quiescent means the stem cells are still in the follicle and are not making new hair. Autophagy, in the release's own definition, is the process by which cells recycle damaged parts so they can keep working. The company says preliminary data indicate that restoring that recycling reactivates the dormant stem cells and supports new growth. Cell assays, tests in cells in a dish, showed increased lactate production. Lactate is a chemical cells make as they break down sugar. The release treats that increase as a reason to keep studying whether the effect can last. Those lines are the company's. The page does not print how large the lactate increase was.

The mouse result, kept as the company's preclinical claim. In a DHT-induced mouse model of androgenetic alopecia, a candidate achieved a hair-growth score comparable to 5% topical minoxidil at one-hundredth of the comparator's concentration. DHT, dihydrotestosterone, is the hormone the model uses to produce that pattern of hair loss in mice. Minoxidil is a drug already sold as a liquid for the scalp. Five percent is a common strength of that liquid. One-hundredth of 5% is 0.05%. That 0.05% figure is arithmetic on the release's two numbers. The release prints one-hundredth. It does not print 0.05%. Comparable means the company says the hair-growth scores lined up. It does not mean the candidate is minoxidil. The release also says other preclinical observations suggest potential activity against inflammation in the follicle. A follicle is the pocket of skin a hair grows from. This is the company's account of its own animal work. The page does not describe a human trial. It does not say an outside lab repeated the mouse test. It does not print the hair-growth score as a number.

How XtalPi says it designed the molecule. The release describes a repeating loop of AI-guided molecular design, physics-based free-energy perturbation, and robotic synthesis and testing. Free-energy perturbation, often shortened to FEP, is a physics calculation that estimates how tightly a molecule holds its target. Robotic synthesis means machines make each new version and run the tests, so the team can try many. The release says that loop was used to improve potency and selectivity along with skin penetration, safety, and whether the molecule can be turned into a topical product. Potency is how strong the effect is at a given amount. Selectivity is whether the molecule hits the intended target rather than others. Skin penetration is whether the medicine actually gets into the scalp. Those lines are XtalPi's. The page does not name the molecule. It does not print a binding score.

What is already on sale, and what that sale is not. XtalPi says it has commercialized two AI-developed ingredients for topical scalp care: Remeanagen, a small molecule, and AquaKine, a peptide. A small molecule is a compact chemical, the usual form of a cream or a pill. A peptide is a short chain of the building blocks proteins are made of. Both have received International Nomenclature of Cosmetic Ingredients names. That list, often shortened to INCI, is the standard naming system for ingredients in cosmetics. The release says the names were granted. It does not print the INCI strings. Groland, a consumer brand XtalPi incubated, has completed U.S. cosmetic facility registration and product listing for a formula that combines the two ingredients. A cosmetic listing registers a beauty product. It is not permission to sell a prescription drug. The release says Groland products are on sale, and that a Groland scalp serum previously ranked No. 1 on Tmall's new-product chart for anti-hair-loss scalp serums. Tmall is a large online store in China. No. 1 is the company's statement. The page does not print the week of that chart, and it does not print a sales total.

What the drug team says it is measuring now. Building on the scalp-care work, XtalPi says it redesigned and optimized candidates for drug development. The team is evaluating how the drug is supposed to work, whether a larger dose does more, how the body absorbs and clears it, how wide the safety margin is, and whether it suits long-term use. Pharmacokinetics is the name for that absorb-and-clear question. A safety margin is the gap between a useful dose and a dose that causes harm. The release says those checks are meant to meet the requirements of a prescription medicine. They describe work underway. They are not results from a study in people.

The market number, and whose number it is. The release cites Global Market Insights. That firm estimates the global androgenetic alopecia treatment market at US$3.4 billion in 2026 and projects it will reach US$6.8 billion by 2035. From 3.4 to 6.8 is a doubling over those years. The doubling is arithmetic on the two figures in the release. The estimate is Global Market Insights', as XtalPi cites it. It is not XtalPi's revenue. The page does not print the company's sales, a valuation, or a funding amount.

The drugs the release sets this against. Topical minoxidil and oral finasteride, the page says, remain the main standards of care for pattern hair loss. They help some patients. Responses vary. People have to keep using them to hold the benefit. The release says topical minoxidil often irritates the scalp, and oral finasteride carries labeled risks of sexual side effects. A labeled risk is a side effect printed on the drug's official label. The company uses that contrast to argue for new mechanisms that reactivate dormant follicle stem cells. Those sentences are XtalPi's account of the existing drugs. The page does not describe a human study of its candidate against finasteride.

Other programs named in the same release, kept separate from this one. XtalPi says the hair-loss step follows other work, including a U.S. IND submission for a gut-restricted treatment for chronic intestinal pain, and the nomination of an HIV preclinical candidate. Gut-restricted means the drug is meant to act in the intestine rather than travel through the rest of the body. A preclinical candidate is a molecule chosen for the studies that come before human testing. Those are different programs. The release does not say the hair-loss candidate has an IND on file. It does not give a date for a hair-loss filing.

Who the about box says XtalPi is. XtalPi Holdings Limited, ticker 2228.HK, was founded in 2015 by three physicists from the Massachusetts Institute of Technology. The box calls it a research and development platform powered by quantum physics, artificial intelligence, and robotics. It says the company combines calculations that start from physics, AI algorithms, high-performance cloud computing, and automated lab systems, and that it provides those tools to companies in pharmaceuticals, materials science, agricultural technology, energy, new chemicals, and cosmetics. The release does not print a price for the hair-loss candidate.

The picture is a magenta card with a white panel and the XtalPi logo, a blue and purple X. A HEALTH label sits on the card. The lines say the hair-loss drug program is entering IND-enabling studies, and they name a lactate–autophagy pathway and an AI and robotics discovery platform. The frame does not print a calendar date. It is not a photograph of a scalp, a clinic, or a mouse. It does not show a score from the mouse study.

In plain terms, XtalPi said on Friday, from Cambridge and Shenzhen, that a topical hair-loss candidate has moved into the studies a company runs before it can ask the FDA to test the drug in people. The company says the molecule is meant to restart dormant hair-follicle stem cells through a lactate and autophagy pathway, and that cells in a dish made more lactate. In mice, it says, one candidate matched 5% minoxidil on a hair-growth score at one-hundredth the concentration. That comparison is the company's animal result. It is not a human trial. XtalPi says it used AI design, a physics binding calculation, and lab robots to tune the molecule for use on the skin. Two earlier AI ingredients, Remeanagen and AquaKine, are already sold in cosmetics under the Groland brand. Cosmetic registration is not a drug approval. The $3.4 billion and $6.8 billion lines are a Global Market Insights estimate, cited by the company. The release does not give an IND filing date, a human result, a valuation, or a funding figure.