Biotechnology · global
China AI Drug-Design Deals Heat Up as Patent Cliff Pushes Multinational Drugmakers Toward New Allies
As patents on blockbuster drugs expire one after another, global drugmakers are more actively turning to Chinese AI drug-design companies for candidate molecules and R&D efficiency. But behind the dealmaking surge, scientific validation, clinical risk, and geopolitical scrutiny remain at the same table.
The timeline for drug development is being tightened again by commercial pressure. For multinational drugmakers that hold blockbuster drugs but are about to face patent expirations, the next drug candidate cannot merely arrive early; it must also have a sufficiently clear biological rationale. That is the backdrop to the recent pursuit of Chinese AI drug-design companies: what they are selling is not just algorithmic imagination, but a pipeline that may move molecules into preclinical and clinical development more quickly.
According to the South China Morning Post, outbound licensing deals between Chinese biotechnology companies and global large pharmaceutical companies increased sharply in the first five months of 2026. Linda Shu, head of China healthcare research at HSBC, noted in a research note that outbound licensing deals signed by Chinese biotech companies with top multinational drugmakers reached US$75 billion in value during the first five months; overall licensing deals totaled 169, worth about US$93 billion, up 87% from the previous year. These figures show a tide of capital and R&D resources moving across borders at the same time.
One of the latest examples is an agreement between AI drug-design company Metis TechBio and Boulevard Bio. According to Hong Kong Stock Exchange filings, Boulevard Bio obtained global development, manufacturing, and commercialization rights to Metis candidate drug MTS-128; Metis may receive a US$20 million upfront payment and is eligible for up to US$1.6 billion in development, regulatory, and commercial milestone payments, as well as future sales profit-sharing. MTS-128 is described as an AI-driven next-generation anticancer drug candidate, but currently public information remains insufficient to assess its target, indications, clinical stage, or the degree of validation already completed.
The specific role of AI here is usually not to replace pharmacology, but to help screen molecules from vast chemical and biological datasets, predict binding ability, tune drug properties, and eliminate unsuitable candidates earlier. If data quality, model assumptions, and experimental feedback can form a virtuous cycle, it may indeed shorten the early discovery stage; the report also cited industry views that AI tools may be able to compress parts of the drug-design timeline from years to about 18 months. However, this kind of time reduction mostly occurs in early R&D and does not mean clinical success rates have been rewritten.
The truly difficult hurdles remain in humans. This is especially true for cancer drugs: tumor heterogeneity, resistance mechanisms, the immune microenvironment, and patient stratification can all cause a molecule that looks promising in models and cell experiments to lose its shine in clinical trials. Therefore, licensing amounts reflect drugmakers’ bets on platforms, pipelines, and speed, and cannot yet be interpreted as proof of efficacy. For serious biomedical readers, the more important information to come will be target mechanisms, animal data, toxicology results, clinical trial design, and reproducible biomarkers.
This wave of deals is also overshadowed by U.S. scrutiny of Chinese biotech and technology companies. For multinational drugmakers, Chinese companies offer speed, cost advantages, and candidate assets; for regulators, data flows, supply-chain dependence, intellectual property, and national security may become intertwined as new issues. AI drug R&D is especially sensitive because it touches biological data, model training, cloud computing, and cross-border licensing at the same time.
As a result, the rise of Chinese AI drug-design companies looks more like a redivision of labor in the global pharmaceutical industry ahead of the patent cliff than a simple technological victory. It may allow more drug candidates to be seen more quickly, and it may also push deal valuations ahead of experimental evidence. The market has already put a price on it; science and regulation will only gradually provide the answers next.