Reading Iambic's S-1: An AI Drug IPO, Risk Factors First

Iambic Therapeutics, a San Diego company that designs drugs with artificial intelligence, set the terms of its initial public offering on October 8, 2026. Its amended registration statement, the S-1/A filed with the Securities and Exchange Commission, offers 9,375,000 shares at an expected price between $15 and $17, on the Nasdaq Global Select Market under the symbol IAM. As of October 14, the offering had not priced. This is a reading of the document, not a view on the shares: nothing here is investment advice, and a preliminary prospectus can still change.

What Iambic says it is

The company describes its approach as molecular superintelligence: large proprietary AI models joined to automated chemistry and biology labs, meant to cover drug discovery from the first hit to clinical candidate. It lists partners including AbbVie, Takeda, Lundbeck, NVIDIA, Revolution Medicines, Jazz Pharmaceuticals and Bayer.

The pipeline is short and early. The most advanced drug, IAM1363, is an oral HER2 inhibitor designed to enter the brain, in an open label phase 1/1b basket trial in patients with advanced HER2 altered solid tumors. Iambic says a registrational trial could start as early as 2027, subject to FDA feedback. A second program, IAM217, aimed at a protein called KIF18A, had an investigational new drug application submitted on September 30, 2026. A third, IAM-C1, a CDK2 and CDK4 inhibitor, is preclinical.

The numbers in the filing

Iambic lost $47.9 million in 2024, $77.3 million in 2025 and $50.1 million in the first half of 2026. Its accumulated deficit was $245.3 million on June 30, 2026. Collaboration revenue was $12.8 million in the first half of 2026, up from $3.9 million a year earlier, coming from the Lundbeck, Revolution Medicines and Takeda agreements.

Cash used in operations was $24.9 million in the first half, far below the loss, mainly because partners paid $33.2 million upfront, booked as deferred revenue. That is money for work still to be done, not profit. Since its founding, the company says it has raised about $461.8 million.

At the $16 midpoint, net proceeds would be about $135.0 million, or $155.9 million if underwriters buy their extra allotment. Iambic says the proceeds and existing cash should fund operations into or through the fourth quarter of 2028. Two existing or new investors, ARK Investment Management and Duquesne Family Office, have indicated interest in buying up to $60.0 million of shares; the filing states that such indications are not binding.

The risk factors, in the company's own words

The most useful pages of any S-1 are the risk summary, because lawyers make the company say plainly what could go wrong. Iambic's list includes:

• a limited operating history, significant losses, and the possibility that it may never be profitable;
• a need for substantial additional capital that may not be available;
• dependence on a small number of partners, including AbbVie, Takeda and Lundbeck, for a substantial share of revenue, and partners' discretion over whether to advance partnered programs;
• an approach to drug discovery that integrates AI with automated experimentation and is, in its words, novel and unproven;
• interim and top line clinical data that may change as more data arrive;
• reliance on third party data, computing hardware, contract research organizations and manufacturers;
• competition from drug companies and other AI drug discovery firms with more resources;
• immediate and substantial dilution for buyers in the offering.

The dilution point has a concrete cause in the filing. Convertible notes will turn into 4,949,237 shares at the offering, at a discount to the public price. Iambic is also an emerging growth company, which lets it file reduced disclosures for several years.

How to read it

Three questions help with any prospectus like this one. First, what has reached patients? Here, one drug in an early trial with no efficacy data presented as a controlled result. Second, where does revenue come from? From partners paying for research, which can stop. Third, what is the money for? The use of proceeds section answers: finishing the IAM1363 phase 1, starting phase 2 trials meant to support possible accelerated approval, then a phase 3, plus early trials for the two other programs and continued platform spending.

For a similar exercise on a larger company, see our reading of Nvidia's 10-Q commitments: the filing usually says more than the press coverage, and says it more carefully.

I find the candor of this document more informative than its ambition. A company that calls its own method unproven, in a filing it must sign, is telling readers where the evidence stands: an interesting platform, a few partnerships that pay for research, and one molecule that has not yet shown in a controlled trial that it works. Whether AI changes drug discovery will be decided in phase 2 data years from now, not on listing day.