OpenAI Startup Fund leads US$153 million round for Biossil, Toronto biotech becomes unicorn at US$1 billion

The Globe and Mail reported on September 28, 2026 that OpenAI has led a US$153-million venture capital financing of Biossil Inc., a Toronto startup that uses artificial intelligence to give failed drug candidates a second chance at becoming approved therapies, valuing the company at US$1 billion. The AI giant, best known for ChatGPT, supplied the majority of the round through its OpenAI Startup Fund, according to the report. The deal makes Biossil the newest Canadian unicorn.

Biossil was founded in 2023 by Anthony Mouchantaf, a lawyer turned entrepreneur who previously headed Royal Bank of Canada's venture capital investment strategy, and Alexander Mosa, who earned a doctor of medicine degree and a PhD in virology and trained at the University Health Network. The company built a software platform that uses OpenAI's large language models to uncover promising molecules from the pharmaceutical discard pile, buying or licensing them from original creators to skip years of early stage work.

The company's research programs reach across several disease areas. BetaKit reported on September 28, 2026 that some of Biossil's research trials target sickle cell disease, idiopathic pulmonary fibrosis, glioblastoma, breast cancer and Alzheimer's disease. NeoTeo reported on September 29, 2026 that Biossil is sponsoring a senicapoc trial for sickle cell disease, with recruitment underway at SickKids, the Centre hospitalier de l'Université de Montréal and McGill University, and that Johnson and Johnson had previously abandoned development of the drug.

The financing arrives during a dense stretch of Canadian artificial intelligence dealmaking. The Canadian AI Newsletter reported on September 29, 2026 that Melbourne based AI medical scribe Heidi raised US$340 million on September 22, that Montreal's Axya raised a C$17-million Series A led by McRock Capital with Yamaha Motor Ventures, and that Astrus, based in Kitchener-Waterloo and Toronto, raised a US$12-million Series A led by Caffeinated Capital for analog chip layout automation, with Khosla Ventures and Garage Capital participating.

Key Facts

The new financing values Biossil at US$1 billion and is an all-equity deal, with a small portion paid to early employees, according to The Globe and Mail. Founders Fund, which led a previous US$22-million financing in 2024, participated, alongside existing investors Quiet Capital, Modern Capital, Golden Ventures and Panache Ventures, plus Duke University's endowment fund and Abu Dhabi's sovereign wealth fund, the Abu Dhabi Investment Council.

OpenAI co-led Biossil's previous venture capital financing a year ago with Peter Thiel's Founders Fund, when the company raised US$43 million, and Janet Bannister's Staircase Ventures led a $3.7-million seed financing in 2023. The all-equity funding will be used to expand Biossil's clinical development capacity and to buy more drug candidates for its project pipeline, BetaKit reported on September 28, 2026. The round is worth about $216 million in Canadian dollars.

The transaction was not publicly announced, and The Globe and Mail attributed its details to a source familiar with the matter. Accounts of the chief executive's response differ. NeoTeo reported on September 29, 2026 that Anthony Mouchantaf declined to comment and that the terms were attributed to an unnamed person familiar with the deal, while BetaKit reported on September 28, 2026 that Mouchantaf confirmed the details to BetaKit in an email. The Canadian AI Newsletter, writing on September 29, 2026, said Mouchantaf confirmed the round in a LinkedIn post.

Biossil disclosed in an August filing with Corporations Canada that it created a new class of 5.37 million preferred shares valued at US$28.69 each, triple last year's share value. As of September 28, 2026, the company had bought or licensed 12 molecules, NeoTeo reported. The approach focuses on repurposing: revisiting a drug candidate developed for one purpose to see whether it might warrant further study, for example because it helped a particular patient subgroup or because the original trial had design or outcome problems.

Senicapoc trials for glioblastoma were described as launching or planned in Germany and Denmark, with Heidelberg University Hospital and the German Cancer Research Center named for the German trial and Aarhus University for the Danish trial. BetaKit reported that Biossil exited stealth this past April with US$70 million in funding, while the Canadian AI Newsletter put the figure at more than US$68 million raised.

Analysis

What this really means is that OpenAI is no longer only selling access to its models. It is financing the companies that convert those models into products in regulated, capital hungry industries, and it is doing so as an owner. Biossil runs on OpenAI large language models, and the OpenAI Startup Fund supplied the majority of a US$153-million round that values the startup at US$1 billion. The strategic logic is legible: if language models genuinely shorten the search for viable molecules, the payoff lands inside the customer, and a fund that holds equity captures part of it.

The valuation itself deserves scrutiny. Biossil's disclosed asset base as of September 28, 2026 is a pipeline of 12 molecules, plus a senicapoc program in sickle cell disease that is still recruiting patients at SickKids, the Centre hospitalier de l'Université de Montréal and McGill University. Nothing in the reporting describes an approved product or a completed late stage trial, and clinical development is where drug programs most often fail. The August filing with Corporations Canada is the clearest marker of how fast investors re-rated the company: 5.37 million preferred shares at US$28.69 each, triple last year's share value.

Set beside the rest of the Canadian artificial intelligence market, the round looks like a bet on a specific thesis rather than a rising tide. Heidi, an AI medical scribe, raised US$340 million on September 22 with Blackbird leading a US$100-million Series C at a US$900-million valuation, plus US$240 million in revenue-based financing from General Catalyst. Astrus raised US$12 million for chip layout automation. The bigger picture here is that Canadian AI capital is spreading across software, semiconductors, health administration and now drug development, and the OpenAI Startup Fund is choosing which of those layers it wants to own.

Drug repurposing is an unusually attractive target for that kind of capital. A molecule that already cleared early safety work and failed for reasons of trial design, dosing or patient selection is cheaper to revisit than a molecule starting from scratch, which is why Biossil buys or licenses candidates rather than discovering them. The risk is that the buyer must be right about why the original sponsor walked away, and Johnson and Johnson's decision to abandon senicapoc is exactly the kind of judgement that will be tested in the sickle cell and glioblastoma trials now underway or planned in Canada, Germany and Denmark.

Why It Matters

Canada has produced relatively few privately held companies worth US$1 billion, and Biossil joins that short list on the strength of a round that The Globe and Mail reported on September 28, 2026. The investor base behind it is unusually international for a Toronto startup: Founders Fund, Quiet Capital, Modern Capital, Golden Ventures, Panache Ventures, Duke University's endowment fund and the Abu Dhabi Investment Council. That mix matters because it shows Canadian health and AI companies can attract both American venture capital and sovereign money at a unicorn valuation.

The clinical stakes are concrete. The programs named in the reporting cover sickle cell disease, idiopathic pulmonary fibrosis, glioblastoma, breast cancer and Alzheimer's disease, conditions where patients have limited options and where a revived molecule could reach trials faster than a brand new chemical entity. Senicapoc is the near term test: it was abandoned by a large pharmaceutical company, and it is now being studied at Canadian academic hospitals with recruitment already open.

The deal also lands against a contradictory backdrop for OpenAI in Canada. The Canadian AI Newsletter reported on September 29, 2026 that British Columbia sued OpenAI over the Tumbler Ridge shooting, that Cohere signed deals with TD and Bell Cyber, and that Shopify became the checkout for Meta's AI agent. The same newsletter noted that Ottawa put $162 million into AI work placements and that Ontario gave the Vector Institute $30 million. An investment in a Toronto biotech sits inside that wider policy and regulatory conversation.

Next Up

Biossil has said where the money goes next: more clinical development capacity and more selective acquisitions of drug candidates, according to BetaKit. Senicapoc trials for glioblastoma are launching or planned in Germany at Heidelberg University Hospital and the German Cancer Research Center, and in Denmark at Aarhus University. Mouchantaf said in a LinkedIn post that the company will have more to say in the coming weeks and months, adding that biomedical insights emerging from AI need to be translated into the physical world to benefit patients.

The open questions are the ones the reporting leaves unresolved. The financing was not announced by the company, and the confirmed details rest on a source familiar with the matter in one account and on the chief executive in others. Watch for a formal announcement, for the next molecules added to the pipeline, and for whether OpenAI's fund keeps leading rounds for startups built on its own models.