#6 China Scholar Insights: The Global Governance of Artificial Intelligence
AI technology is a double-edged sword. International efforts are needed to set safety standards and balance innovation and security needs.
Welcome to the sixth edition of China Scholar Insights!
China Scholar Insights is a feature which aimed at providing you with the latest analysis on issues that Chinese scholars and strategic communities are focusing on. We will carefully select commentary articles and highlight key points. Questions or criticisms can be directed to sunchenghao@tsinghua.edu.cn.
I am SUN Chenghao, a fellow with the Center for International Security and Strategy (CISS) at Tsinghua University. ChinAffairs+ is a newsletter that shares Chinese academic articles focused on topics such as China’s foreign policy, China-U.S. relations, China-European relations, and more. This newsletter was co-founded by me and my research assistant, ZHANG Xueyu.
Chinese Scholars on The Global Governance of Artificial Intelligence
Background
The rapid development of AI has made global governance essential. The AI Safety Summit and the Responsible AI in the Military Domain Summit achieved progress, such as the Bletchley Declaration promoting cooperation and the Seoul Summit creating the International AI Safety Institutes Network. The AI Action Summit in February 2025 saw 61 countries sign the Declaration on Inclusive and Sustainable AI. However, the UK and US’s refusal to sign underscores governance challenges driven by technological competition, highlighting that effective global AI governance remains difficult.
Summary
AI is advancing rapidly, but faces challenges like declining returns, centralization, and social imbalances. China needs innovation-friendly regulations, better integration of research and industry, and mechanisms to foster diverse innovation.
As a double-edged sword, AI technology reshaping national security, warfare, and governance, but also posing risks like system vulnerabilities and blurred governance boundaries. International efforts are required to set safety standards and balance innovation with security.
Currently, global AI governance faces issues like uncertain AI development, coordination challenges, and geopolitical tensions. A balanced framework that ensures inclusivity and effectiveness is essential. China has contributed the Global AI Governance Initiative, promoting agile governance and US-China cooperation, and the international community must collaborate on technical standards and security research.
Insights
XIAO Qian: The World Needs Inclusive and Open International Governance of AI — Observations and Reflections from the French AI Action Summit
The AI Action Summit issued the “Statement on Inclusive and Sustainable AI”, advancing France’s goals of promoting an open, diverse AI governance model, enhancing France’s global AI influence, closing the gap with China and the US, and developing its domestic AI ecosystem. However, the US and the UK refused to sign the statement, and the UK claimed that it was not in line with its national interests.
The Divergence in AI Governance Concepts May Impact US-EU Cooperation
The US-UK-led summits prioritized AI security within a Western-centric framework, while France’s summit emphasized openness, commercial interests, and broader themes like public interest and global governance. During the summit, JD Vance’s speech criticized the EU’s AI regulatory policy, indicating the Trump administration’s proposition to relax the regulation of AI companies and also reflecting that the divergence in AI governance concepts between the US and Europe is constantly expanding, which may hinder future cooperation.
China Has Always Played a Constructive and Stable Role
From the Global AI Governance Initiative and the Resolution on Enhancing International Cooperation in Capacity-Building of Artificial Intelligence to the AI Capacity-Building Action Plan for Good and for All, China has always been promoting global joint actions to address the development of AI. During this summit, the China AI Development and Security Research Network held an official side event themed “Progress in Artificial Intelligence Technology and Its Applications”, which was highly recognized by various countries. China has become an important force playing a constructive, stable and positive role in this summit.
Future Task: Balancing the Representativeness and Effectiveness of AI Global Governance Framework
The world needs an inclusive and open international governance of AI. This requires countries to cultivate similar concepts and cognition in AI theory and practice through continuous international exchanges and cooperation. In addition, it is urgent to build a global AI governance framework in which all countries can participate. How to balance representativeness and effectiveness is a key issue that the AI international governance framework needs to consider and an unfinished task left by the French summit for future summits.
XUE Lan: Risks of Artificial Intelligence and Strategies for Global Governance
Intrinsic Characteristics and Misapplications: Risks of AI Systems
AI systems have inherent vulnerabilities, including security flaws, opaque decisions, and ethical risks like privacy breaches and algorithmic bias. They can be misused for deepfakes, misinformation, and cyberattacks, threatening social order and critical infrastructure. Meanwhile, military applications may escalate conflicts, while technological competition risks triggering an arms race.
Currently, the international community is deepening its understanding of AI-related risks. Entities ranging from the United Nations to national governments and private sectors are actively engaging in international cooperation to explore governance pathways. However, effectively mitigating risks posed by AI technologies remains a formidable challenge.
From Technology to International Politics: Challenges to Global Governance
AI governance faces challenges due to the unpredictable trajectory of general AI, lagging legal frameworks, and complex international coordination. U.S.-China technological decoupling further hinders global AI collaboration, with geopolitical tensions obstructing cooperation on dual-use technologies. As decoupling extends to other scientific domains, exchanges on commercial technologies and cutting-edge innovations face growing barriers.
Institutional Building and International Collaboration: Addressing Global Governance Challenges
To mitigate AI risks, mechanisms should address ethical concerns at the source, while governments adopt agile governance. Given the complexity of global AI governance, hierarchical and collaborative strategies are needed. AI risk management should be treated as a global public good, avoiding exclusionary approaches. Despite geopolitical tensions, selective U.S.-China tech cooperation remains crucial.
After DeepSeek’s Emergence: How China Can Build a New AI Revolution and Foster an Innovative Environment
As China’s DeepSeek disrupts the AI industry, Trump calls it a “wake-up call” for U.S. tech. His policy focuses on boosting domestic innovation while tightening external blockades. In this context, China must seriously assess the institutional conditions necessary for sustained innovation in cutting-edge technologies.
The Reasons for the Development of American Technology: Institutional Conditions Supporting Diverse Exploration
Open environment: In an open institutional environment, new ideas and initiatives are given the opportunity to be tested in the U.S., helping to sustain its vitality and creativity in cutting-edge technologies.
Industry-Research Collaboration: The strong collaboration between industry and research is pivotal to ensuring continued progress in the U.S. AI sector. Additionally, the self-iteration and competition among a few dominant American tech giants form a crucial foundation for advancing frontier technologies.
Financial Support for Innovation: Federal and state governments provide various financial resources, including venture capital, public welfare funds, and corporate investments, driving innovation and competition crucial for the U.S. to maintain its global AI leadership.
Analysis of the Issues in China’s AI Development Policy: Lack of Innovation
Regulatory Approach: China’s “heavy regulatory” mindset needs a shift toward “innovation-friendly regulation” by respecting market mechanisms and implementing targeted deregulation in key sectors.
Innovation Model: China has long viewed technological innovation as a linear process of “basic research→applied research→commercialization of scientific achievements,” which artificially separates the research activities from industrial development. Although the Chinese government has initiated relevant guidance and development plans, the coordination between scientific research and industrial policies remains weak.
Funding and Industry Support: Unlike the U.S. NAIRI, which integrates foundational and applied research across sectors, China’s funding channels for scientific and technological innovation remain relatively narrow. Research is primarily concentrated in universities and research institutes and the industry lacks support for innovative talent.
The Special Importance of creating an environment for Open Innovation in China and Policy Suggestions
In recent years, the U.S. has attracted many Chinese AI talents, exacerbating brain drain from China. Trump’s AI executive order aims to reduce federal oversight and foster a flexible regulatory environment. China needs institutional reforms and greater openness.
Policy suggestions include aligning with emerging international rules and exploring new regulatory frameworks, such as relaxing the internet firewall. Additionally, funding methods should be diversified by linking public and social funds to support AI development. Collaborative platforms between research institutes and enterprises should be developed, and talent development should be innovated, particularly by encouraging young students to study advanced technologies. Lastly, the development of underdeveloped regions can be promoted through paired assistance initiatives.
LU Chuanying:Artificial Intelligence Reshaping National Security Paradigms and Logic
AI, a strategic and dual-use technology, is reshaping national security, much like past technologies such as nuclear weapons, space tech, and the internet. Its versatility spans both civilian and military applications, raising concerns about pan-securitization in national security.
Artificial Intelligence Empowering National Security Systems
AI-driven warfare is evolving from labor-intensive to intelligent systems, improving efficiency, precision, and reducing casualties. Countries with advanced AI dominate information warfare, using data analytics to identify vulnerabilities and inform decisions. In intelligence, AI processes vast data streams to detect anomalies and provide timely alerts. The private sector’s AI innovations challenge government control over national security, introducing both vitality and potential conflicts between corporate and national interests.
The Profound Impact of Artificial Intelligence Technology Security on National Security
The integration of AI into national security has highlighted issues such as diversified risks, increased system vulnerabilities, blurred governance boundaries, and greater uncertainty in threat perception. AI introduces new threats like cyberattacks and economic espionage, while its “black box” nature complicates threat detection. National security strategies must be flexible and adaptable. AI’s complexity, data reliance, and vulnerabilities require enhanced security standards, innovation, and data protection. Additionally, AI’s widespread use reshapes governance, making traditional frameworks inadequate for addressing cross-domain risks.
National Security Governance System in the Artificial Intelligence Era
In the AI era, national security governance must evolve with a comprehensive AI security framework, focusing on technical, regulatory, and social resilience. This requires synergy between standards, policies, and strategies, alongside cross-departmental collaboration. Governments, industry, and research institutions must work together to address AI risks and prioritize early warning and response mechanisms, balancing technological progress with security.
LIANG Zheng: China is shaping AI governance mechanism
The Urgency of Global AI Governance
As AI nears the “technological singularity”, its potential is vast, but its complexities present global challenges. A coordinated global governance framework is urgently needed, as AI transcends borders, creating risks and opportunities that no single country can manage alone. International collaboration is essential to mitigate these risks effectively.
International Collaboration to Govern AI for a Safer Future
As AI advances rapidly and large models become more autonomous, governance structures struggle to keep up. To ensure AI benefits humanity, action is necessary to prevent harm. In 2023, global summits on AI safety were held, and countries like the US, Canada, the UK, Singapore, and Japan established AI safety research institutes. This reflects a growing global consensus on strengthing AI governance and cooperation, laying the foundation for a more unified and secure global system.
Challenges in Global AI Governance
Global AI governance faces several challenges, including a lack of openness, fairness, and effectiveness. Western countries have created “exclusive circles”, limiting broader participation and undermining discussions. Many countries in the Global South struggle with limited infrastructure, resources, and talent, increasing their vulnerability to AI’s social and economic impacts. Additionally, the governance system is fragmented, with countries competing to influence global AI standards.
China’s Unique Path in Global AI Governance
China’s contribution to global AI governance emphasizes “Eastern Wisdom”, advocating for open, fair, and effective governance. By promoting an inclusive platform within the UN framework, China supports multilateralism, international cooperation, and fair participation. It stresses balancing development with security, aiming to create a responsive and adaptable governance system in the fast-evolving AI landscape, prioritizing collective security and balanced development.
Conclusion
Artificial intelligence (AI) technology has a powerful influence on national security. International governance is required to balance the needs of innovation and national security. However, the current international governance of AI is faced with multiple challenges. How to build a global AI governance framework has become an important issue for the international community. This necessitates that the international community transcend geopolitics, consider AI as a global public good, engage in coordinated efforts, and enhance the processes of standard-setting, the development of risk response plans, as well as the improvement of the governance system. Through the establishment of a sound global governance system, the world will eventually develop inclusive and sustainable artificial intelligence for the benefit of humanity and the planet.
Writers and Editors for Today’s Newsletter:
Writers:
BAI Xuhan, WANG Jiaying, GAO Liangyu, HU Lingzhi, and HNIN Lei Lei Wai, Tsinghua University
Editors:
SUN Chenghao, Tsinghua University
ZHANG Xueyu, Nankai University