China recognises 16 companies as AI x Biomanufacturing pioneers
13 of the 16 projects are in biological design to drive strategic autonomy from US tools.
Key points
In Aug 2025, China unveiled its first cohort of 16 AI-enabled biomanufacturing best-practice cases. (full list at end of article)
13 of the 16 projects apply AI to biological design rather than to downstream manufacturing.
The reasons for these selections are probably threefold:
Strategic autonomy: developing alternatives to U.S.-dominated biological design tools.
Technical opportunity: Current AI is well-suited to biological design, making it a domain where high-quality projects are available today.
Long-term impact: Advancing biomanufacturing requies better organisms
AI in biomanufacturing
Biomanufacturing takes a microbe such as E. coli or yeast, or alternatively cells from mammals, engineered to produce a particular target product, and grows them in bioreactors fed with sugars, amino acids, carbon dioxide and nitrogen. These low-value inputs are taken up by the biological system and converted into wide-range or higher-value molecules that can be harvested, purified, and sold. The end products range widely from medicines to monosodium glutamate to plastics.
The issue is that even these ‘simple’ organisms have biological systems that are made up of a huge number of components that interact with each other in different ways.
This complexity is where biology diverges from conventional industries. For a long time biomanufacturing has lagged behind because the fundamental science hasn’t been solved. Effectively, companies were building LEGO without an instruction booklet.
AI is writing the instruction booklet.
At its most advanced, AI could govern all stages of the biomanufacturing value chain. Design of novel proteins or metabolic pathways could be modelled in silico by AI [that means on a computer], and then validated experimentally using robotics-enabled automated labs in a closed-loop fashion. Production of the target compound could then be scaled in intelligent bioreactors [with few employees] and adjustments made to the culture medium or process parameters in order to optimise the yield. At the commercial scale, continuous operation could be achieved using real-time AI-driven quality control.
The strategic importance of AI in biotech
The strategic importance of biomanufacturing has two dimensions to it; capability and economic security. AI will be a key enabling technology.
On capability, nations want their scientists to have access to state-of-the-art tools to enable world-class research. AI is increasingly embedded in biological research, so leading research means leading AI-driven tools or using AI driven tools better than others. From a Chinese perpsective, the US is the undispited global leader in biological design tools.
On economic security, government investment in industrial biotech is motivated by the potential to re-shore manufacturing and to thereby secure supply chains for critical chemical inputs. AI will be used at every step. The economic security advantage of biomanufacturing is that, theoreatically at least, it can make a tremendously wide range of products.
The focus of the August announcement was AI-powered biological design tools. Not manufacturing
In August 2025, China’s Ministry of Industry and Information Technology announced the first cohort of sixteen AI-enabled application best practice cases in biomanufacturing, as part of a broader national mission to incorporate AI into industry. The announcement was worded such that this seems the first of multiple tranches.
The original call for applicants identified eight major focus areas that fall into different buckets along the biomanufacturing value chain:
Design and construction of high-performance proteins (高性能蛋白质元件的设计及构建): modelling of protein structure and design of protein sequence to optimise traits such as thermostability, catalytic activity, substrate specificity, and target affinity etc.
Analysis and optimisation of gene regulation mechanisms (调控机制的解析及优化): exploration of the regulatory mechanisms of gene expression for the development of predictive models, enabling dynamic control of genes and products, and construction of gene circuits.
Design and optimization of metabolic pathways (代谢通路的设计及优化): discovery and design of new metabolic pathways to open up new biosynthetic pathways or increase yield of known metabolites
Construction and optimization of cell factories (细胞工厂的构建及优化): model the correlation between genotype and industrial indicators (e.g. stress tolerance, yield) and use high-throughput gene editing technology to improve the adaptability of cell lines to industrial contexts
Design and optimization of culture medium formulation (培养基配方的设计及优化): use IOT-based monitoring to obtain real time fermentation process data. Combine with metabolomics to find the optimal ratio of substrates, key nutrients, and trace elements in the culture medium.
Intelligent control of biological reaction processes (生物反应过程的智能控制): Utilise intelligent sensing and control technologies to test control strategies such as precise feeding, variable speed stirring, and aeration adjustment, important process parameters (e.g. temperature, pH, dissolved oxygen) in bioreactors. Use digital twins and other means to simulate bioreactor processes in order to accelerate the process of fermentation condition optimization and process scale-up.
Intelligent detection and quality control of biomanufactured products (生物制造产品的智能检测和质量控制): achieve closed-loop quality control of biomanufacturing processes by automatically analysing data from sensors and feed quality information back to the production control system, thereby enabling real-time dynamic optimization of process parameters.
Other cases (其他案例): other AI-enabled cases that empower the biomanufacturing industry. What this refers to exactly is unclear from the text, but I imagine it could include AI models implicated in feedstock processing or energy management.
This is a major scope of applications. And yet, the current awardees are concentrated across four of these focus areas.
9 in AI for protein design
3 for construction and optimisation of cell factories (a cell factory is the whole engineered organism or cell that is a “factory” to produce the desired output)
3 for intelligent control of biological reaction processes (this is precise of the biomanufacturing process, so manufacturing )
1 for design and optimisation of metabolic pathways.
13 of these 16 tools (the protein design, metabolic pathways, cell factories) are AI-powered biological design tools to be used when designing the organism. The drivers of this could be concern over US control of of the tools, a recognition that better organisms are needed for China to suceed in the industry, or maybe they were simply the best applicants for the original call-out (it could be that simple).
As computational biological design tools have proliferated, many have become more narrow focusing on subsets of proteins or metabolic pathways for which they have been trained. That is case for some of the projects with proteins being designed for relatively narrow applications (such as Vanillin production). Others look to be a little more broad ranging.
Some of the datasets for the Chinese tools are possibly (we do not know for sure) partially built on open-source US tools like Alphafold.
As far as we can tell, these companies were already working on these areas, so the announcement is not driving the awardees into new territories.
What does this actually mean?
This announcement is a vote of confidence in these companies and the work they are doing, rather than a forward-looking public investment in the form of a grant or major loan. It probably is meant to signal to the system (provincial governments, universities, governments, financiers) that these companies and similar style projects should be supported. It probably also signals a first step into bigger AI x bio policy directions.
We can also question the extent to which this will galvanise the Chinese biomanufacturing industry. While the Chinese government is sending a signal with announcements like these, market competition on its own will tend to drive companies to adopt these technologies if they are useful at the right price point. They might see other opportunities. At the very least the potential government support may figure in their calculations.
The Chinese government has a broad library of strategies for biomanufacturing that focus on bioreactors, scale-up, feedstocks, and end products. No other nation has anything comparable to this level of planning.
 | A guest post by
|