Five-year plan stands up Panshi/ScienceOne model, intelligent-scientist systems, a scientific corpus, data centers and a research ‘capability map’
The Chinese Academy of Sciences released its development plan Monday for the 15th Five-Year Plan period, 2026 through 2030. Integrated circuits, artificial intelligence, quantum technology, brain-computer interfaces and controlled nuclear fusion sit on the short list of important priority fields.
Life sciences and biotechnology are on that list too. The academy called the document its overall roadmap and its construction blueprint for the next five years. Strategic high-technology work covers optoelectronics, space, information, ocean, energy and materials. Sustainable-development work covers agriculture, Earth and resources, and the atmosphere, ocean and ecological environment. Basic frontier and interdisciplinary work covers mathematics, physics, chemistry, biology, astronomy and geoscience.
CAS is treating AI as more than another research topic. A dedicated chapter frames it as a lever for changing how science gets done and how labs get run. The academy said it will use existing strength in mathematics, computer science, control science and scientific data. New infrastructure is supposed to include the Panshi science foundation model — carried in English reports as the ScienceOne Omni foundation model — along with intelligent-scientist systems, a scientific corpus and scientific data centers. The stated point is coordinated support, not a pile of disconnected tools.
National major science infrastructure and national field observation stations are slated for digital, networked and intelligent upgrades. Data work is bundled with that: collect more of it, stitch it together, unify standards, harden security, then push fusion and reuse into key studies and major projects.
Management is getting its own stack. CAS will run a special task called Intelligent Chinese Academy of Sciences and build an integrated intelligent research-management platform. Three centers sit underneath. One is meant to allocate innovation resources across the academy. One is meant to command and dispatch science and technology activity. One is meant to pull research-management information into a single collaborative layer. Officials also want a research “capability map” that is comprehensive, kept current and drawn with precise profiles of teams and labs.
That map is the inventory. The model stack is the engine.
Geography is part of the plan, not an afterthought. Beijing and the Beijing-Tianjin-Hebei region, Shanghai and the Yangtze River Delta, and the Guangdong-Hong Kong-Macao Greater Bay Area are treated as strategic highlands. Hefei, Chengdu-Chongqing, Wuhan and Xi’an are strategic pivots. Other key regions are nodes. The aim, as written, is systematic support for the country’s major regional strategies.
Industry is written in as a pull, not only a destination. The Chinese readout stresses deeper fusion of scientific innovation and industrial innovation and the cultivation of new quality productive forces. English accounts describe a two-sided approach: supply-side advances in research paired with demand-side pull from industry, so results move into production. The focus, again, is important fields, key regions and major projects.
Talent is tied to the same timetable. The plan keeps education, science and technology, and talent development on one track and calls for building national strategic talent capacity. Other reform items listed for a world-class research institution include a national science and technology think tank, changes in how research organizations are managed, transfer and application of results, and international science and technology exchanges.
No officials were quoted by name in the academy notice or the Xinhua accounts. The Chinese notice, posted on the CAS site via China News Service, described the plan as an active embrace of AI for discovery and for administration.
Executives and developers watching compute, silicon, scientific software and lab data will find the named build-outs in one place: Panshi/ScienceOne, the intelligent-scientist systems, the corpus, the data centers, the three management centers, the capability map, and the highland-pivot-node grid. Those are the pieces CAS said it will stand up between 2026 and 2030.