Chinese biotech out licensing deals hit US$93B in the first five months of 2026, an 87% jump year over year, as global drugmakers hunt replacements for top sellers like Keytruda, Eliquis, and Opdivo that lose US patent protection by 2029.
The reason Chinese AI-designed drug candidates are landing in Big Pharma pipelines at an unusually early stage is sitting in four years of US patent expirations: 2026 to 2029. Over that window, the patents on several of the world's top-selling medicines, including Merck's Keytruda, Bristol-Myers Squibb's Opdivo, the Pfizer and Bristol-Myers Squibb blood thinner Eliquis, and Eli Lilly's Trulicity, will expire, opening the field to generics and biosimilars and forcing their owners to refill multibillion-dollar revenue holes with new assets (Genetic Engineering & Biotechnology News, Optum, Proclinical).
That structural pressure is showing up as deal flow. The South China Morning Post, citing an HSBC research note by China healthcare research head Linda Shu, reported that Chinese biotech out-licensing deals with the world's top pharma companies reached US$75 billion in the first five months of 2026, up from essentially zero before 2020. Over the same period, Chinese biotechs struck 169 licensing deals worth a combined US$93 billion, an 87% jump year over year (SCMP, citing HSBC's Linda Shu). Those numbers come from a single bank note and have not been independently confirmed against a third-party deal database, so they should be read as an HSBC estimate rather than a settled fact.
The clearest single proof point for what the trend looks like in practice is a deal METiS TechBio, a Hangzhou-based AI drug-design company, signed with Boulevard Bio, a Deerfield-backed biotech. METiS handed over global development, manufacturing, and commercialization rights to MTS-128, a next-generation trispecific T-cell engager, a synthetic antibody designed to grab three different targets on cancer cells at once, and built by METiS's AI platform (METiS press release on PRNewswire, AllSci coverage). The headline value is up to US$1.6 billion in development, regulatory, and commercial milestone payments, with a US$20 million upfront and tiered royalties on any future sales. The catch: MTS-128 is preclinical, with no human trials yet, meaning the US$1.6 billion is contingent and only realized if the candidate clears the usual gauntlet of clinical, regulatory, and commercial gates (BioPharma APAC coverage).
The deal's filing on the Hong Kong Stock Exchange, where METiS is listed, lays out the same terms: preclinical asset, modest upfront, large contingent milestones, global rights (METiS HKEX filing, 30 June 2026). For a global pharma buyer, paying for a candidate that has not yet entered the clinic is unusual, and the fact that an AI-designed Chinese asset cleared that bar illustrates how thin the late-stage pipeline looks for the companies staring at the cliff.
The demand side, in other words, is not a fad. The 2026 to 2029 expirations on top blockbusters leave a multi-hundred-billion-dollar revenue gap that the existing internal pipelines of large pharma companies are not expected to fill (Optum analysis). Chinese AI platforms, which can compress preclinical candidate generation from years to months, have become a logical external source of replacement assets.
That supply line is now running into a parallel constraint. SCMP reports that US scrutiny of Chinese biotech deals is rising in step with the volume, with regulators in Washington growing uneasy about cross-border licensing of drug candidates tied to emerging AI and data infrastructure (SCMP). The SCMP account does not name the specific US policy instruments, such as a possible expansion of the Committee on Foreign Investment in the United States review process or a new biotech-focused provision, but it describes a tightening regulatory backdrop.
That is the tension the deal flow now sits in. The demand for new preclinical assets is structural, driven by a multi-year loss of exclusivity on drugs that already define Big Pharma's revenue base, and the supply of those assets is concentrated in Chinese biotech companies that use AI-driven design platforms. The earliest signal of whether the US can slow the flow will be how regulators treat the next deal of METiS's size. If a US-headquartered partner signs one, the pattern is durable. If every US buyer is forced to step back, the licensing pipeline reroutes to Europe and Japan, and the US patient base waits longer for the replacements.