Eli Lilly's $2.75 billion partnership with Silicon Integrated: Behind the push, R&D costs can't suppress expansion impulses, financial pressure is urgent, seeking breakthroughs for survival

The reporter Guo Yilin and Na, from chinatimes.net.cn, Beijing

Eli Lilly’s rumored acquisition of Exscientia faces a reversal—are we about to enter the era of “self-funding” AI drug discovery?

On March 29, rumors that Eli Lilly would acquire Exscientia quickly spread through the market. According to Reuters, Eli Lilly, a major U.S. pharmaceutical company, had reached an agreement to acquire Exscientia and had already obtained approval from the U.S. Federal Trade Commission (FTC). Shortly afterward, Ning Feng, CEO and Chief Scientific Officer of Exscientia, responded jointly, saying: “We’ll wait for our official update early tomorrow morning.”

Against this backdrop, outside observers became even more certain about the acquisition rumor. However, later that evening, the deal took a turn: Exscientia issued an announcement stating that it had reached a global exclusive licensing and research collaboration with Eli Lilly. Under the agreement, Exscientia will receive a $115 million upfront payment and, after achieving development, regulatory, and commercialization milestones, will receive cumulative potential payments of up to approximately $2.75 billion. The company will also enjoy tiered royalty payments based on future sales.

In a clarification, Alex Zhavoronkov, founder and CEO of Exscientia, said the likely reason the market generated acquisition rumors was the FTC’s antitrust review. “Because the transaction value is huge and reaches the mandatory filing threshold, a submission to the FTC is required and approval must be obtained. It is exactly this regulatory step that the market misread as an M&A transaction.”

This “acquisition” suspense ultimately ended as a “licensing” arrangement. But the astronomical figure of $2.75 billion is enough to make this transaction itself a milestone for the AI drug discovery industry. In response, Zhang Siyuan, a special research fellow at Suning Bank, told reporters from The Huaxia Times, “This misunderstanding is more like an accidental pressure test that precisely examined the market’s sensitive nerves and proved the real commercial value in the AI drug discovery field. Currently, AI drug discovery is still in a critical transition period from ‘telling stories’ to ‘delivering answer sheets.’ Take Exscientia as an example: its business is highly dependent on a revenue model centered on drug development and payments from a small number of major customers. This is not an isolated case in the AI drug discovery industry; its ‘high risk, high reward’ characteristics run through the entire process. When there is a lack of diversified licensing arrangements and new revenue sources, the company’s sensitivity to macro cycles and changes in drug policy increases, limiting its resilience to risks.”

The details of the misinterpretation of the transaction

On the evening of March 29, Exscientia released its 2025 annual performance results. At the same time, it officially disclosed an external licensing agreement and a global R&D collaboration with Eli Lilly. Under the agreement’s terms, Eli Lilly will pay Exscientia a $115 million upfront payment, plus milestone payments for development, regulatory, and commercialization, as well as tiered profit-sharing from future sales. The total transaction value can reach up to $2.75 billion.

At this scale, in today’s global biopharma BD market, it is already a typical “top-tier pricing” deal. It is understood that even the U.S. Federal Trade Commission triggered an antitrust review. According to an official FTC release, the transaction filing threshold for 2026 is about $130 million, and the upfront payment under the parties’ collaboration already touched that red line. The information showing review approval was officially updated on March 26, 2026.

What’s interesting is that this regulatory action was misread by the market as an M&A transaction. On March 31, Alex Zhavoronkov, founder and CEO of Exscientia, clarified that the company is not likely to be acquired by a pharma company and still intends to remain independent. “We are not selling now—at least not during the recent period. We believe the valuation of Exscientia is still too low and can’t reflect the true value.”

From the transaction structure, Eli Lilly will obtain global exclusive development, manufacturing, and commercialization rights for a potential “best-in-class” oral innovative drug from Exscientia that is in the preclinical stage and targets specific indications. Meanwhile, the two parties will also jointly develop multiple targets selected by Eli Lilly. By combining Exscientia’s Pharma.AI platform with Eli Lilly’s clinical development capabilities, they aim to accelerate the discovery and advancement of innovative candidate drugs.

The underlying asset of this collaboration is widely speculated to be closely related to the GLP-1 field. At the performance briefing on March 30, Exscientia’s presentation materials disclosed a pipeline in which global rights had been licensed to an “undisclosed partner.” The target is GLP-1R, used to treat metabolic diseases. Eli Lilly’s “new drug king,” tirzepatide, which has annual sales of $36.5 billion, has GLP-1R as one of its core mechanisms of action. And this GLP-1R-target product from Exscientia has completed lead compound optimization and is now preparing for an investigational study submission. For Eli Lilly, this is tantamount to buying a “research acceleration coupon.” Andrew Adams, Vice President of Molecular Discovery at Eli Lilly, said directly that Exscientia’s AI capabilities are “a powerful complement” to Eli Lilly’s clinical development capabilities.

In fact, this super collaboration was not accomplished overnight. It went through a three-stage leap: from “testing the technology,” to “strategic endorsement,” and then to “binding interests.” The parties’ connection can be traced back to 2023, when Eli Lilly introduced Exscientia’s Chemistry42 platform to launch a software licensing collaboration. In November 2025, the parties further reached a strategic AI drug R&D collaboration totaling more than $100 million. After that, Eli Lilly also became the first company to back the Hong Kong-listed Biotech IPO with its corporate identity, and as a cornerstone investor it subscribed for $5 million worth of shares in Exscientia. Regarding this, Ren Feng said: “In 2023, Eli Lilly subscribed for three years of Chemistry42 software services at once. Starting with a small collaboration and building trust step by step is what gave Exscientia the opportunity to sign a super collaboration worth $2.75 billion.”

Sheng Meng, Executive Director of Shangxiang Capital, told reporters from The Huaxia Times that in today’s business markets, when enterprises allocate resources, they focus more on the effectiveness of investment and the input-output ratio. Therefore, companies would not raise a dairy cow just to get a cup of milk. The cost the acquisition team must bear is far greater than acquiring R&D results that have already passed the risk period.

Two ledgers behind the $2.75 billion

To understand why Exscientia could quote a “top-tier price” behind this collaboration, you need to examine it from the positions of both transaction parties. For Eli Lilly, the R&D costs cannot curb the impulse to expand. For Exscientia, financial pressure urgently needs a way out.

On December 30, 2025, after four rounds of filings, Exscientia listed on the Hong Kong Stock Exchange. According to publicly available financial information, Exscientia’s 2025 revenue was about $56.24 million, down 34.5% year over year. Net losses were $352 million, and losses widened significantly year over year, mainly driven by a one-time non-cash fair value loss from the conversion of preferred shares totaling $297 million.

It is worth noting that the change in revenue structure is even more intriguing. Revenue from drug discovery jumped from $3.14M in 2024 to $24.95M in 2025, raising its share of total revenue from 3.7% to 44.4%. However, pipeline development revenue fell sharply from $76.59M in 2024 to $23.89M in 2025. Between this increase and decrease, the “cliff-like fragility” of Exscientia’s revenue model is exposed: it is highly dependent on milestone payments from a small number of major customers. In the first half of 2025, with the lack of Exelixis-related milestone payment support, revenue dropped from $59.69M in the same period of the prior year to $27.46M, a decline of 54%.

In the full-year 2025 revenue structure, revenue from drug discovery was $24.95M; revenue from pipeline development was $23.89M; revenue from software solutions was $4.91M; and revenue from other discovery activities was $2.49M. By the end of 2025, the company’s cash and bank balances were $393M. Although ammunition was supplemented through the E-round financing and the IPO, based on the current “cash burn” pace, operations still face ongoing pressure.

For Eli Lilly, although it delivered a strong set of financial results in 2025, the other side of its high-speed growth is rapidly expanding R&D expenses. The company’s total revenue in 2025 reached $65.2 billion, up 45% year over year. Its core products Mounjaro and Zepbound contributed more than $13.0 billion in revenue in the fourth quarter, up 91% year over year. R&D expenses were $13.34B, up 21% year over year, accounting for more than 20% of total net revenue.

It is also worth noting that as early as the first quarter of 2025, Eli Lilly lowered its full-year adjusted earnings per share outlook due to higher R&D spending and market price adjustments—from the prior $22.50–$24.00 range down to $20.78–$22.28. In the first half of 2025, R&D expenses were $6.07 billion, up 16% year over year; for the full year under non-GAAP standards, R&D spending increased 26% year over year.

An even more noteworthy signal is that Eli Lilly expects 2026 revenue to be $80 billion to $83 billion, up about 25% year over year. Exponential growth in scale means that relying on the traditional “linear expansion” model of internal R&D is no longer sufficient to match the company’s growth ambitions. Against the industry backdrop of an approaching patent cliff, there are about $236 billion worth of drugs globally facing a patent cliff, and Eli Lilly urgently needs an efficient mechanism to continuously “produce” candidate molecules.

This explains why, since 2025, Eli Lilly has rolled out new AI initiatives almost every month. From March 2025, when it collaborated with Nvidia to build what the industry calls the strongest AI supercomputer for pharma, to November, when it reached a $345 million bispecific antibody development collaboration with Ailux under the Jing Tai Technology group, to signing an R&D collaboration agreement worth more than $100 million with Exscientia—Eli Lilly is using real money to build “infrastructure” for AI-driven drug discovery.

In response, Andrew Adams, Vice President of Molecular Discovery at Eli Lilly, said: “Exscientia’s capabilities in AI-driven drug discovery complement Eli Lilly’s experience in clinical development and disease research. This is expected to help both sides discover new mechanisms of action and accelerate the development of candidate drugs.”

When will AI drug discovery achieve “self-funding”?

In fact, when a company whose R&D investment accounts for one-fifth of revenue still wants to speed up, an AI drug discovery company becomes the “accelerator.”

Against this backdrop, the strategic significance of this Eli Lilly transaction is self-evident. The $115 million upfront payment is nearly twice Exscientia’s full-year revenue, directly injecting substantial cash flow; the overall $2.75 billion package provides a predictable channel of milestone revenue over the coming years. In this regard, a research report from Southwest Securities estimates that Exscientia’s revenue in 2026–2028 will reach $272 million, $323 million, and $390 million, respectively.

In an analysis of this reversal, Shi Tianyi, an analyst from the medical and healthcare division of Hejun Consulting, told reporters from The Huaxia Times that the core significance of the reversal is not the transaction value itself, but the fact that it tears open the long-standing “fog” around AI drug discovery. Eli Lilly’s willingness to pay “clinical-stage pricing” even in the preclinical stage is more of a premium acquisition of R&D certainty. And this industry is moving from an “Earning PPT” phase into an “Earning in clinical trials and competing on orders” realization period.

In recent years, AI drug discovery has been one of the most questioned sectors for “telling stories.” In this “vicious cycle,” fundraising has been hot and valuations have been inflated, yet commercialization paths remain unclear. Each time someone asks, “Do you really get patients to use the drugs?” the answers have always been ambiguous.

Although Eli Lilly’s $2.75 billion collaboration breaks this deadlock, the challenges facing AI drug discovery companies have not disappeared. Exscientia’s 2025 annual report shows R&D expenses of $81.40 million, and an adjusted net loss of $43.80 million. Software solutions revenue was only $4.91 million, accounting for less than 9% of total revenue. This indicates that despite the BD collaboration being hot, the underlying “software-as-a-service” business model for AI drug discovery has not yet worked in practice; the industry still heavily relies on the heavy-asset model of “selling pipelines and selling molecules.”

责任编辑：姜雨晴 主编：陈岩鹏

A massive amount of information and precise interpretation are available in the Sina Finance APP