Laboratory Profile

The Beijing Key Laboratory of Artificial Intelligence Safety and Superalignment addresses the safety risks and governance imperatives posed by frontier AI technologies. The Laboratory focuses on achieving breakthroughs in AI safety and superalignment technologies, exploring solutions for the safe and controllable development of AI systems, realizing effective oversight, guidance, and risk control for frontier AI models and superintelligence, and conducting demonstration applications on frontier AI models. In doing so, it aims to erect cutting-edge safety and governance guardrails for the innovative development of artificial general intelligence (AGI) and to guide superintelligence toward harmonious coexistence with humanity.

The Laboratory is established under the leadership of the Institute of Automation, Chinese Academy of Sciences, with Peking University and Beijing Normal University as co-founding partners. It brings together an interdisciplinary research team spanning artificial intelligence, cognitive psychology, brain science, ethics, safety and governance, and systems science. The Laboratory aims to leverage and consolidate Beijing's strengths in interdisciplinary research, target key scientific questions and practical challenges in AI safety and superalignment, reshape and optimize the existing scientific and technological landscape of AI safety and superalignment, and construct a "Beijing Approach" for the safe and controllable development of AI and superintelligence.

Scientific Research Objectives

  • To investigate and construct AI ethics and safety models and frameworks

  • To develop new theories and models for safe AI that integrate passive risk prevention with proactive design

  • To explore and develop new theories and technologies for superalignment that combine external supervision alignment with internal mechanism alignment

  • To construct a "Beijing Approach" to superalignment that ensures AI safety and controllability

Development Goals

  • To become an internationally recognized, cutting-edge interdisciplinary research base for AI safety and superalignment

  • To consolidate Beijing's strengths in interdisciplinary R&D and provide forward-looking leadership in frontier research on AI safety and superalignment

  • To reshape and optimize the existing scientific theories and technological landscape of superalignment

  • To attract and cultivate innovative talent in AI safety and superalignment for AGI and superintelligence

Laboratory Logo

图片

Laboratory Structure

The Beijing Key Laboratory of Artificial Intelligence Safety and Superalignment is established under the leadership of the Institute of Automation, Chinese Academy of Sciences, with Peking University and Beijing Normal University as co-founding partners. It comprises over 50 researchers, forming an interdisciplinary research team spanning artificial intelligence, cognitive psychology, brain science, ethics, safety and governance, and systems science.

图片

图片

图片

图片4

图片

图片

图片

图片

图片

图片

On March 29, 2025, the "Beijing Collaborative Innovation Matrix for AI Safety and Governance" was unveiled at the Zhongguancun Forum. As a member of this Innovation Matrix, the Beijing Key Laboratory of Artificial Intelligence Safety and Superalignment, together with six other Beijing Key Laboratories—including the Beijing Key Laboratory of AI Safety and Standards—jointly advances collaborative innovation in AI safety and governance in Beijing.

图片

图片

Laboratory English Name

Beijing Key Laboratory of Artificial Intelligence Safety and Superalignment

Laboratory Website

https://beijing.ai-safety-and-superalignment.cn

图片


Beijing Key Laboratory of Artificial Intelligence Safety and Superalignment

The Beijing Key Laboratory of Artificial Intelligence Safety and Superalignment addresses the safety risks and governance imperatives posed by frontier AI technologies. The Laboratory focuses on achieving breakthroughs in AI safety and superalignment technologies, exploring solutions for the safe and controllable development of AI systems, realizing effective oversight, guidance, and risk control for frontier AI models and superintelligence, and conducting demonstration applications on frontier AI models. In doing so, it aims to erect cutting-edge safety and governance guardrails for the innovative development of artificial general intelligence (AGI) and to guide superintelligence toward harmonious coexistence with humanity.

The Laboratory is established under the leadership of the Institute of Automation, Chinese Academy of Sciences, with Peking University and Beijing Normal University as co-founding partners. It brings together an interdisciplinary research team spanning artificial intelligence, cognitive psychology, ethics, safety and governance. The Laboratory aims to leverage and consolidate Beijing's strengths in interdisciplinary research, target key scientific questions and practical challenges in AI safety and superalignment, reshape and optimize the existing scientific and technological landscape of AI safety and superalignment, and construct a "Beijing Approach" for the safe and controllable development of AI and superintelligence.

© 2024-2026 Home | Beijing Normal University