The developer of the world’s first end-to-end AI drug has projected a big leap in half-year earnings, driven by lucrative collaborations with multinationals

image credit: Bamboo Works

Key Takeaways:

  • Insilico expects to power into the black after cashing in on drug partnerships this year with Eli Lilly, Takeda and SK Biopharmaceuticals
  • Its flagship drug, which was discovered and developed using AI, has entered Phase Three trials but substantive results are not expected before late 2027

For years, artificial intelligence has promised to speed up drug discovery, but the research has rapidly burned through cash while struggling to get clinical or commercial traction.

That could be starting to change. A Chinese biotech that uses AI to discover and design new drugs for cancer and chronic diseases has just flagged up an impending surge in earnings, backed by a raft of deals with pharmaceutical giants.

The profit alert from Insilico Medicine (3696.HK) comes just six months after it turned heads with Hong Kong’s biggest IPO of 2025, riding on investor hopes for AI-assisted medical breakthroughs.

Insilico projected a swing into profit for the first half of this year, with six-month revenue nearly double the firm’s whole turnover for 2025.

The company forecast revenue of between $102.5 million and $106.5 million, a surge of more than 270% from the same period a year earlier, turning a loss of $19.2 million in the first half of 2025 into a projected net profit this time ranging from $33.5 million to $39.5 million. Adjusted non-IFRS net profit was estimated to be between $45.5 million and $51.5 million.

The main driver was a host of licensing and collaboration agreements with leading names in the global drug industry, focusing on finding new treatments for cancer, lung disease and autoimmune conditions. According to the DXY Insight database, which tracks global pharmaceutical markets, Insilico has sealed 14 deals so far in 2026 with a total potential value exceeding $7 billion. These include an agreement worth up to $888 million in January with French drug firm Servier, a $2.75 billion global R&D collaboration with Eli Lilly (NYSE:LLY) in March, a $2.5 billion pact with SK Biopharmaceuticals in June and a $600 million global strategic collaboration with Takeda (NYSE:TAK) in early July.

It is worth noting that the headline figures include milestone payments and royalties on potential sales, with no certainty that the money will be received in full. Moreover, earnings generated by these business development deals are inherently volatile. The agreements can deliver a quick earnings boost, but subsequent revenue streams depend on clinical progress and commercial viability.

The bumpy pattern is already visible in Insilico’s earnings. Last year revenue fell to $56.24 million from $85.83 million in 2024, in the absence of any big licensing deals, while its annual net loss swelled to $352 million from just under $17.10 million.

End-to-end AI drug

The AI drug discovery industry generates revenue in three main ways: selling software platforms, providing outsourced R&D services and licensing internally developed drug candidates. But Insilico is not content with being a mere provider of AI technology. The company is also using its flagship AI platform to develop its own drugs that could be independently launched on the market.

By the middle of this year, Insilico had identified 31 preclinical candidates across high-barrier, high-potential therapeutic areas including oncology and immunology, as well as disorders of the metabolism and central nervous system. Of these, 13 had received regulatory clearance for testing as investigational new drugs. Of 10 programs in clinical development, four are being advanced independently and six are progressing through collaborative clinical studies with pharmaceutical partners.

Insilico’s most advanced candidate has just entered late-stage trials, billed as the world’s first drug in which the biological target and the therapeutic compound were discovered and designed using generative AI. Rentosertib (ISM001-055) began Phase Three trials on July 8 as a treatment for idiopathic pulmonary fibrosis, a chronic lung disease. U.S. regulators have designated it an "orphan drug", entitled to financial incentives as a potential treatment for a rare condition, while China has included the drug in its breakthrough therapy program. But despite the excitement, there are underlying concerns. Part of its Phase Two testing enrolled only 71 patients, while the U.S. Food and Drug Administration previously put a temporary hold on a trial of the drug. While the testing later resumed, the episode showed that regulators apply extremely rigorous safety protocols to molecules identified and designed using AI.

Assuming a 52-week course of treatment, Rentosertib’s Phase Three trial is expected to wrap up at the end of June next year. Allowing extra time to process the data, the key results are unlikely to emerge before the end of 2027. That means the project, initiated in 2019, could take a decade to proceed through testing to regulatory approval and commercial launch. If AI accelerates the early stages of drug discovery but fails to improve success rates during the far more expensive and failure-prone clinical phase, its value to the pharmaceutical industry could be substantially diminished.

Insilico commands a market value of about HK$26.5 billion ($3.38 billion), compared with approximately HK$31 billion for fellow AI drug leader XtalPi Holdings (2228.HK), whose revenue is less dependent on licensing deals. Insilico has committed heavily to its internally developed pipeline. The future of its share price depends on whether the flagship drug proves effective in clinical tests, and whether the firm can forge new partnerships to keep generating cash.

To subscribe to Bamboo Works weekly free newsletter, click here

Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.