Another traditional industry has officially "fallen in love with AI".

On March 29, local time, American pharmaceutical giant Eli Lilly announced a strategic cooperation with Hong Kong-listed AI pharmaceutical company Insilico Medicine (Insilico)—a preliminary payment of 115 million dollars,along with subsequent milestone payments, with a potential total value reaching 2.75 billion dollars, plus future sales tiered royalties.

This figure made the whole industry pause for a second—has the "GPT moment" of AI pharmaceuticals arrived?

Before today, AI drug discovery was more like a story that has been repeated but never concluded. Startups raising funds, big corporations entering the fray, academia providing endorsements… but whenever someone asked, "Have you really come up with a drug that patients use?," the answer was always ambiguous.

This payment from Eli Lilly represents a clear change in stance.

In fact, Eli Lilly is also accelerating its "AI transformation".

At the J.P. Morgan Healthcare Conference in March, Eli Lilly just announced the establishment of a 1 billion dollar joint innovation AI laboratory with NVIDIA, specifically to address long-standing challenges in the pharmaceutical field. In the same month, NVIDIA also reached a collaboration with Novo Nordisk to accelerate drug discovery using the Gefion sovereign AI supercomputer.

But that's not all. In February, Takeda Pharmaceutical signed a collaboration with AI company Iambic Therapeutics worth over 1.7 billion dollars, aimed at using AI to find new drugs for cancer and other diseases. On March 27, Quotient Sciences and Intrepid Labs just announced a years-long collaboration to introduce the machine learning model ANDROMEDA into early drug development.

This is not an isolated case; this is a collective bet. As 2026 begins, the pharmaceutical industry is witnessing a flurry of major AI platform deals, with Lilly, Sanofi, Novo Nordisk, Bayer…almost every top pharmaceutical company is rushing to sign contracts.

Industry analysts' figures are more straightforward: the AI drug discovery market will be worth about 2.9 billion dollars by 2025, expected to reach 5.1 billion dollars by 2026, and exceed 13.4 billion dollars by 2035.

However, just because there is a lot of hot money doesn’t mean the problems have vanished.

Insilico Medicine is not a new company.

This company was founded by Chinese scientists and is headquartered in Hong Kong. It is currently one of the very few companies that has truly advanced AI-generated drug molecules into clinical stages. Its core technology involves using generative AI to directly "design" entirely new molecular structures, rather than just screening existing compound libraries—this represents a fundamental difference in the technical approach.

Traditional drug discovery typically follows this process: identify targets → repeatedly screen through millions of known compounds → find candidate molecules → lengthy optimization. This process averages over a decade and costs billions of dollars.

Insilico's approach is somewhat like "drawing a key from scratch"—telling AI directly, "this is what the lock looks like, design a key that can open the lock." Its end-to-end platform Pharma.AI encompasses three core stages: target discovery, molecule generation, and clinical outcome prediction. The company claims to have compressed some drug discovery cycles to under 18 months.

Insilico's CEO stated directly: "The only company stronger than us in AI is Eli Lilly; there is no other company."

This sounds very bold. However, Lilly's willingness to offer this price is in itself a form of endorsement.

What Lilly needs is not just an AI tool but a "production line" that can continuously generate drug candidate molecules.

However, understanding this deal requires understanding its structure.

The 115 million dollar upfront payment is real money that happens today, the kind that shows up in the account. But the remaining 2.6 billion dollars are "milestone payments"—these will only be fulfilled gradually after Insilico's AI model successfully produces validated targets and candidate molecules that enter human trials, and even complete clinical trials.

This can be seen as a structure that expresses strategic intent through "potential upper limits of value", using "initial investment" to test actual capabilities. For Lilly, the balance sheet will not immediately bear the pressure of 2.75 billion dollars; for Insilico, every milestone payment represents a public report card.

Industry comments on this deal structure are quite uniform: it limits Lilly's direct risk while providing Insilico with the strongest commercial incentive—if you don’t deliver, you won’t get paid.

But this brings us to the problem.

Drugs discovered by AI still need to go through human clinical trials. The failure rate of clinical trials is frustratingly high across the entire pharmaceutical industry—about 90% of candidate drugs do not survive phase two trials. Whether AI-generated molecules can truly break this curse currently lacks sufficient data to answer.

One industry observer's assessment is relatively objective: "The prediction for 2026 is that validation and disappointment will each account for about half. The field has moved from speculation to early clinical validation, but the gap between promise and performance remains large."

Without any drugs ultimately reaching the market or receiving regulatory approval,the entire AI drug discovery field is actually still in a very lengthy "proof of concept phase".

No matter how many large contracts exist, they cannot change this outcome.

When Eli Lilly decided to embed AI into its core strategy, rather than just a laboratory budget, its significance has gone beyond the scope of a single business contract.

Looking back, one of the most obvious trends in the AI industry over the past two years has been "vertical penetration"—large model capabilities seeping into various specialty fields, from code generation and legal documents to now molecular design. The pharmaceutical industry, known for its conservatism, strict regulations, and lengthy research and development cycles, is theoretically one of the slowest fields for AI penetration.

But the signal is very clear now: the most conservative money has begun to flow.

Lilly's choice also has a deeper industrial logic. Semaglutide brought Novo Nordisk to the top of global pharmaceuticals, and Lilly's Tirzepatide has also achieved great success in the weight loss market. This "GLP-1 war" has taught all pharmaceutical companies one thing: whoever finds the next "target" first will win the next decade.

And AI is currently the most likely tool to accelerate this "finding" process.

Insilico's current 2.75 billion dollar valuation is more than just a transaction; it is a ticket to enter—a signal to the entire industry that AI drug discovery has evolved from "research curiosity" into "business reality".

In the next two to three years, if the candidate molecules discovered by AI can demonstrate statistically meaningful advantages in clinical trials, then this "AI drug discovery revolution" can be said to be truly underway. If the molecules entering clinical trials in bulk continue to fail based on traditional probabilities, the industry will face a period of calm, trading enthusiasm will cool, and Insilico will face real tests.

No one knows the outcome.

But Lilly’s current 115 million dollar check represents the most expensive vote so far.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。