The Computer Network Information Center of the Chinese Academy of Sciences (CAS) launched the Panshi AI-for-Science Ecosystem 1.0 on Sept. 30 at the CAS Academic Hall in Beijing.

The initiative combines Panshi OneScience, an AI-enabled research platform that underpins the ecosystem, with the Xinhuo Program, a pilot initiative that supports applications of the platform and development of the broader ecosystem.

Panshi OneScience integrates research data, models and tools to support AI-assisted research. It is designed to connect national scientific facilities and laboratories and to support the research workflow—from hypothesis development and literature review to simulation, experimentation, data analysis and discovery.

The platform also includes capabilities for evaluating and tracing AI-enabled research. Its evaluation framework covers five disciplines: mathematics, physics, chemistry, biology, and Earth and space science. A traceability system for research resources is intended to support verifiable research processes and data governance.

Nearly 100 research institutes, universities and technology companies have joined the Xinhuo Program. Their work spans foundation models for science, research corpora and application development.

The program will give participants access to shared validation environments, computing resources, data sample sets and end-to-end research toolchains. It focuses on interdisciplinary talent development, collaboration among research institutions, universities and industry, and the translation of research results into practical applications.

The initiative is intended to help partners address common research challenges and expand the use of AI in science, according to CAS.