Conference Seminar | New Trends in Global AI Innovation Governance: Key Takeaways from the Recent Seminar

The subsequent open discussion session further stimulated deep exchanges among the attending guests. They not only shared frontier thoughts on the application and governance of AI but also proposed constructive development paths centered on how China, and particularly Shanghai, can play an active role in the global AI governance landscape.

Cai Cuihong, a Researcher at CGAIG and Professor at the Institute of International Studies at Fudan University, analyzed the current landscape and challenges of global AI governance alongside Chinese solutions. Professor Cai noted that AI is becoming a key variable in reshaping the global power structure, and the overall landscape of global AI governance presents characteristics of being institutionalized, multi-centered, and multi-level, gradually forming a multi-polar interactive trend anchored by China, the US, and Europe, with other nations constantly following up. Regarding challenges, she highlighted that AI governance faces multi-layer nesting and fragmentation alongside institutionalization trends; a few developed countries and tech giants monopolize most computing power and data; value divergences make it difficult to unify standards; and governance lags behind rapid technological iteration. In response, Professor Cai suggested progressively building consensus through issue-specific zoning combined with soft law. Regarding China's role, Professor Cai emphasized continuing to exert value leadership, enhancing technical and industrial strength to increase discourse power in rule-making, deepening South-South cooperation to help developing countries share AI benefits, and actively participating in multilateral dialogues with the US and Europe to seek consensus while reserving differences.

Xu Wenwei, a Professor at the Center for Technology Innovation and Strategy at Fudan University, discussed the dilemmas of AI deployment and the necessity of governance from an industrial perspective. Current issues include: first, unrealistically high expectations for AI, which is still in the weak AI stage; second, the need for vigilance against AI hegemony and inequality; third, doubts regarding the explainability and reliability of AI; fourth, the importance of ethical issues, including support for the workforce and corporate norms; and fifth, the urgent need to establish interoperability standards across industries to enable data flow and sharing among various sectors.

Chen Disi, Director of Public Affairs at Alibaba Intelligent Information Business Group, shared the challenges and practices of large internet companies in AI safety governance. Quark is Alibaba's flagship AI application, integrating tools such as AI search, AI writing, AI problem-solving, AI health assistant, AI image generation, AI PPT, and AI resume to create an all-around AI assistant for users' work, study, and life. In terms of governance, Quark has built a strategy, capability, and organizational system covering the entire process to create safe and reliable AI applications: for instance, embedding governance into the full link, using AI to govern AI to enhance safety capabilities, and employing problem-driven security operations for real-time prevention. The interactive service model of large models brings diversification of compliance subjects, and the capability for self-learning makes data hunger a norm. Director Chen suggested that enterprises should proactively improve model safety awareness and risk control capabilities and implement AI labeling to guarantee users' right to know; the industry could share high-quality datasets like malicious samples; and he expressed hope for the speedy release of guidelines on safety liability division among diverse subjects to promote healthy and orderly industry development.

Xu Qi, Deputy Secretary-General of the Shanghai Artificial Intelligence Industry Association, introduced Shanghai's industrial development in the AI field. Shanghai possesses a full-stack technical system ranging from infrastructure, chips, and computing clusters to foundation models and application scenarios, laying a foundation for comprehensive governance. In this context, the key role of the AI Industry Association in linking and empowering the industrial ecosystem should be fully leveraged to promote the development of the AI industry. Subsequently, Deputy Secretary-General Xu shared several thoughts: first, AI ethical governance cannot be separated from innovation, productization, and commercialization across technology, ecology, and industry. Second, standards are crucial, and enterprises should run fast with small steps to drive underlying innovation and external cooperation. Third, he suggested establishing an AI export practice base to provide all-around support for domestic large models and AI-related products in applying and serving overseas markets. Fourth, he proposed leveraging Shanghai to promote the export of open-source and inclusive AI products to the Global South, empowering them with infrastructure and models through open co-construction of international scenarios and deepening cooperation with Global South nations.

Hui Zhibin, Director of the Internet Research Center at the Shanghai Academy of Social Sciences, explained his understanding of AI governance work from macro and strategic perspectives. First, AI governance is a dynamically evolving process where new technologies constantly spawn new governance issues. Second, there is a game of concepts among nations regarding AI governance. From a strategic judgment, China must possess sufficient institutional confidence and strategic focus, adhere to the path of open cooperation, and provide more public goods for AI governance on the international stage. Third, AI governance is a trust issue in the digital world, and China can promote mutual trust and consensus on a broader scale through technical trust. Based on this, Director Hui suggested further thematic research on the development dilemmas and security risks caused by the development of artificial intelligence technology, as well as promoting deep interaction and cooperation between academia and industry.

Zhu Rongsheng, Senior Researcher at Qiyuan Insight Think Tank, argued that promoting the AI governance industry could be an opportunity to alleviate the insufficiency of AI commercial profitability. Shanghai's promotion and development of the AI governance industry can improve technical safety, create economic value, and help shape China’s image as a responsible major power internationally. He believed that the establishment of the Center for Global AI Innovative Governance (CGAIG) is an important step for China to participate responsibly in global governance, which not only helps drive the global AI governance process but also allows China's good practices to go global while introducing advanced international development and governance experiences.

Xiao Yanghua, a Professor at the School of Computer Science and Technology at Fudan University, offered deeper reflections on governance based on the essence of AI development. First, Professor Xiao emphasized the need to clarify concepts and distinguish between problems inherent to human society and governance issues native to AI. Second, this AI revolution is fundamentally different from previous industrial revolutions; AI is beginning to partially replace humans in cognition and decision-making, making it a revolution of the cognitive subject. Third, governance and development are not simply opposites; good governance can protect sustained and healthy technological innovation. Finally, strategic height should be raised, as AI governance is not merely technical governance but concerns the future direction of human civilization. Based on this, Xiao proposed suggestions such as establishing a risk classification governance system for AI, designing governance systems around different stages of AI development, increasing interdisciplinary research efforts, remaining vigilant against potential risks in the process of AI application, and forming a consensus-based AI governance framework.

After the guest speeches, Li Anfang, Director of the Shanghai Philosophy and Social Science Planning Office, provided a concluding review of the seminar. He noted that this seminar achieved a fusion of ideas and a collision of wisdom across government, industry, academia, and research, fully highlighting the critical role of cross-sector integration in addressing AI governance challenges. He emphasized that AI is not only a core direction for global industrial development but also relates to the long-term US-China competitive landscape and China's national strategic layout. Shanghai is actively integrating into this process and taking initiative; it has established multiple important platforms, including the Center for Global AI Innovative Governance (CGAIG), and has preliminarily demonstrated its positive role in overall coordination and leading development.

At the conclusion of the meeting, Zhang Yi, Executive Dean of the Fudan Development Institute, thanked the attending guests for their support on behalf of the organizers. She pointed out that current AI governance still faces a certain degree of fragmentation and lacks overall coordination. In the future, supported by a larger system, deep linkage among academia, industry, and other stakeholders should be promoted to form synergy. The vision for future AI development should be to build consensus, face the future, promote symbiosis, and achieve fusion within the global governance process.