On July 11, four departments including the Beijing Municipal Science and Technology Commission and the Administrative Committee of Zhongguancun Science Park jointly issued the Beijing Action Plan for Accelerating AI-Enabled High-Quality Development of Scientific Research (2025-2027) (hereinafter referred to as the “Action Plan”). It aims to promote in-depth integration of artificial intelligence (AI) and scientific research in Beijing and build a globally influential source of AI innovation.
The Action Plan proposes that significant progress will be made in multiple key areas by 2027. It focuses on four aspects: key technological breakthroughs, infrastructure construction, domain-specific application implementation, and innovation ecosystem development, providing a clear direction for accelerating AI-driven high-quality development of scientific research.
1. Strengthening Key Technological Breakthroughs
The plan focuses on breakthroughs in basic theories of scientific intelligence, exploring cutting-edge methods such as AI-based Schrödinger equation solving and electronic structure calculations to overcome bottlenecks in multi-scale modeling theories. Meanwhile, it will accelerate the development of interdisciplinary general scientific foundation models to achieve unified representation, modeling, analysis, and generation of multi-type, high-dimensional scientific data. It will also develop next-generation intelligent simulation software tools for scientific computing, including multimodal scientific information data collection and annotation tools, to provide efficient and accurate technical support for researchers.
2. Consolidating Infrastructure Construction
In terms of infrastructure, the plan proposes building an open and shared scientific data platform to promote the aggregation and trusted circulation of high-quality scientific data resources and develop multimodal large models for scientific knowledge. It will establish a computing power platform with hardware-software collaboration, develop dedicated computing chips, optimize computing power scheduling mechanisms, and reduce computing power usage costs. Additionally, it will construct intelligent dry-wet closed-loop experimental systems, develop automated experimental devices and intelligent unmanned laboratories, and form an efficient R&D model of “computation guiding experiments and experiments feeding back to computation” to significantly improve research efficiency.
3. Accelerating Domain-Specific Application Implementation
The plan will drive scientific intelligence to assist in exploring laws, solving problems, and simulating phenomena in basic scientific research. In the medical and health field, it will build expert-level databases to accelerate drug R&D and intelligent diagnosis and treatment. In new materials, it will optimize the entire process of material design and preparation. In scientific instruments, it will tackle intelligent equipment to enhance the competitiveness of domestic instruments. In industry, it will promote intelligent upgrading through simulation software and vertical large models, creating a batch of replicable and promotable benchmark application cases.
4. Cultivating a Favorable Innovation Ecosystem
Beijing will build a scientific intelligence public service innovation platform and a research collaboration cloud platform to accelerate the industrialization of research outcomes. It will introduce high-level research teams and cultivate interdisciplinary talents proficient in both AI technology and application. A multi-channel investment and financing service system will be established to guide financial institutions in innovating service models. It will also develop characteristic industrial clusters, strengthen international cooperation and open-source ecosystem development, and form a sound pattern of collaborative development integrating “industry, academia, research, application, and finance.”
To ensure the implementation of the plan, Beijing will strengthen cross-departmental collaboration, innovate project organization methods, enhance supply-demand connection, and rely on expert committees to provide decision-making support. As these measures advance, the in-depth integration of AI and scientific research will bring fundamental changes to Beijing’s research paradigm, accelerate original innovation breakthroughs, drive high-quality industrial development, and help Beijing gain an advantage in global technological competition.
