Lightning Talk #246 AI for Sustainable Development Public Private Sector Roles

Summary

This discussion at the IGF 2025 focused on how artificial intelligence can advance sustainable development and the respective roles of public and private sectors in this process. The session was hosted by Tsinghua University and featured speakers from academia, government, and youth perspectives. Professor Yong Guo opened by emphasizing AI as a strategic technology with strong spillover effects, highlighting Tsinghua University’s commitment to intelligent society governance through their Institute for Intelligent Society Governance established in 2019. He stressed three key areas: talent development with interdisciplinary education, strengthening collaborative mechanisms across sectors, and deepening international cooperation.

Ms. Xuanyun You from China’s Cyberspace Administration outlined the Chinese government’s approach to AI governance, including President Xi Jinping’s Global AI Governance Initiative and China’s proposal for AI capacity building action plans. She detailed China’s comprehensive regulatory framework encompassing AI policy rules, data security laws, and online information governance, emphasizing principles of balancing development with security and implementing inclusive supervision. Youth Ambassador Xin Yi Ding presented the younger generation’s perspective, acknowledging AI’s benefits in areas like fire prediction, autonomous vehicles, and rural economic development, while warning about risks including data monopolies, algorithmic biases, and the dangerous proliferation of deepfakes and misinformation.

Professor Rony Medaglia from Copenhagen Business School provided research-based insights on AI’s dual impact on sustainability, presenting positive examples in efficiency improvements, data governance, and sustainable business models, while cautioning about negative effects including high energy consumption, water usage for server cooling, rebound effects, and embedded biases in AI systems. The discussion concluded that realizing AI’s potential for sustainable development requires careful mitigation of risks while maximizing opportunities through coordinated international efforts.

Keypoints

## Major Discussion Points:

– **AI’s Role in Advancing Sustainable Development Goals**: The discussion emphasized how AI can contribute to achieving the UN’s 2030 Sustainable Development Goals through improved efficiency, data governance, and sustainable business models, with concrete examples like water management systems and pollution monitoring.

– **International Cooperation and Governance Frameworks**: Speakers highlighted the importance of global collaboration in AI governance, including China’s Global AI Governance Initiative and UN resolutions on AI capacity building, particularly focusing on helping developing countries benefit from AI advancements.

– **Balancing AI Opportunities with Risks**: The conversation addressed both the positive potential of AI (enhanced public services, educational access, economic opportunities) and significant risks (misinformation, deepfakes, algorithmic bias, energy consumption, and threats to public trust).

– **Youth Perspective and Generational Responsibility**: A key focus was placed on the younger generation’s unique position as “digital natives” who must navigate AI’s challenges while taking responsibility for ethical AI development and promoting digital literacy across society.

– **Public-Private Sector Collaboration**: The discussion emphasized the need for coordinated efforts between government, industry, academia, and research institutions to develop appropriate regulations, standards, and ethical frameworks for AI development.

## Overall Purpose:

The discussion aimed to explore how artificial intelligence can be leveraged to advance sustainable development while addressing the respective roles of public and private sectors in this process. The session sought to bring together diverse perspectives from academia, government, youth, and international research to identify both opportunities and challenges in using AI for global sustainability goals.

## Overall Tone:

The discussion maintained a consistently optimistic yet cautious tone throughout. Speakers were enthusiastic about AI’s potential to solve global challenges and advance sustainability, but they balanced this optimism with realistic acknowledgment of significant risks and challenges. The tone was collaborative and forward-looking, emphasizing the need for international cooperation and responsible development. There was no notable shift in tone during the conversation – it remained professionally optimistic while being appropriately mindful of the complexities involved in AI governance and sustainable development.

Speakers

– **Cynthia Su** – Deputy Dean of Tsinghua Range Joint Research Institute for Intelligent Society at Tsinghua University, China; Session moderator

– **Yong Guo** – Vice Chair of University Council of Tsinghua University; Professor

– **Xuanyun You** – Associate Deputy Director-General of the Bureau of Law-Based Cyberspace Governance at the Cyberspace Administration of China

– **Xin Yi Ding** – Youth Ambassador of the Institute of Intelligent Society, Governments of Tsinghua University

– **Rony Medaglia** – Professor at Copenhagen Business School (also referred to as “Ron and Anthony Meddalia” in introduction, but appears to be the same person based on context)

Additional speakers:

None identified beyond the provided speakers names list.

# Discussion Report: AI for Sustainable Development – Public and Private Sector Roles

## Executive Summary

This session at the 20th United Nations Internet Governance Forum was hosted by the Tsinghua Range Joint Research Institute for Intelligent Society. The listening talk brought together speakers from Chinese academia and government, international research, and youth advocacy to examine how artificial intelligence can advance sustainable development goals and the roles of public and private sectors in this process.

The discussion featured four main speakers: Professor Yong Guo, Vice Chair of University Council at Tsinghua University; Ms. Xuanyun You, Associate Deputy Director-General of the Bureau of Law-Based Cyberspace Governance at China’s Cyberspace Administration; Youth Ambassador Xin Yi Ding from the Institute of Intelligent Society Governance; and Professor Rony Medaglia from Copenhagen Business School. The conversation addressed both AI’s potential benefits for sustainability and the significant challenges that must be addressed.

## Detailed Discussion Analysis

### Academic and Institutional Framework

Professor Yong Guo opened by referencing President Xi Jinping’s statement that AI “has strong spillover effects that can drive broader progress” as a strategic technology driving scientific revolution and industrial transformation. He highlighted Tsinghua University’s Institute for Intelligent Society Governance, established in 2019 with Professor Su Jun as founding dean, which focuses on fundamental theories and core policy issues of AI integration into society.

Guo outlined three critical areas requiring attention: developing talent through interdisciplinary education that combines technical skills with social responsibility; strengthening collaborative mechanisms across government, industry, academia, and research institutions; and deepening international cooperation through technology exchanges, standard setting, and financial integration to ensure equitable sharing of digital dividends. He emphasized that universities must lead AI development to serve communities and enhance accessibility through interdisciplinary research approaches.

### Government Policy and Regulatory Approach

Ms. Xuanyun You presented China’s comprehensive approach to AI governance, detailing President Xi Jinping’s Global AI Governance Initiative emphasizing people-centered, AI-for-good principles. She referenced specific UN resolutions, including the 17th session resolution on “Seize the Opportunities of Safe, Secure, and Trustworthy AI Systems for Sustainable Development” and the 1978 session resolution on “Enhancing International Cooperation on Capacity Building of AI.”

You outlined China’s regulatory philosophy of “giving equal importance to development and security, innovation and governance,” implementing “inclusive, prudent, and classified and graded supervision of AI.” She detailed China’s legal framework including the “Next Generation AI Development Plan,” “interim measures for the management of generative AI services,” and comprehensive policies covering AI governance, data security, and online information management.

The government’s approach focuses on reducing biases, misinformation, and security threats while building capacity in computing power, data management, and governance structures. You emphasized China’s commitment to international cooperation through AI Capacity Building Action Plans designed to help developing nations strengthen their AI capabilities and participate in the global digital economy.

### Youth Perspective and Digital Literacy

Youth Ambassador Xin Yi Ding opened with direct questions: “Have you ever been tricked by the information online or paused before sharing a viral news thinking it might be AI-generated? These are not just hypothetical questions, they are profound signs of a huge shift.” She demonstrated how AI can create convincing deepfakes that blur lines between reality and fabrication, potentially harming public trust and democratic governance.

Ding highlighted positive AI applications including fire pattern prediction systems, autonomous vehicles with collision avoidance technology, and AI-powered educational platforms. She provided a detailed example of rural economic development: a golden retriever in Zigui County helping sell oranges (Mao Mao fruit) through AI-powered marketing, generating 4 billion yuan in sales by March 2024.

She emphasized that young people bear responsibility in both public and private spheres for AI governance, rejecting “symbolic participations” and demanding substantive involvement in shaping AI development. Ding stressed the need for improved AI literacy to help everyone understand data usage and algorithmic logic while developing critical thinking skills to guard against AI-generated misinformation.

### International Research and Environmental Considerations

Professor Rony Medaglia provided research-based insights on AI’s dual impact on sustainability, identifying three key areas: efficiency improvements, enhanced data governance, and sustainable business model development. He presented concrete examples including Aarhus city using AI sensors in water grids to anticipate usage and reduce waste, Danish drone companies monitoring emissions, and blockchain-AI combinations enabling product provenance tracking for environmental verification.

Medaglia introduced critical data about AI’s environmental costs: “generating one single image with a large language model uses the same amount of CO2 as charging your mobile phone up to 50%.” He detailed how generative AI systems require significant energy resources and create substantial water withdrawal demands for server cooling – “six times the whole country of Denmark for a whole year of water.”

He discussed rebound effects, where efficiency gains can lead to increased resource consumption, using car-sharing applications potentially reducing public transportation use as an example. This highlighted how AI’s benefits can have unintended consequences requiring systems-level thinking rather than assuming efficiency automatically equals sustainability.

## Key Areas of Agreement

All speakers acknowledged AI’s transformative potential for sustainable development across sectors including disaster prediction, resource management, urban governance, and business model innovation. There was consensus on the necessity of multi-stakeholder collaboration involving government, industry, academia, and international institutions, with particular attention to ensuring developing countries can participate in AI advancements.

Speakers agreed that AI poses serious risks requiring attention alongside its benefits, including bias and discrimination, misinformation and deepfakes, energy consumption and environmental impact, and potential threats to public trust and governance systems.

## Different Emphases and Approaches

While agreeing on fundamental principles, speakers emphasized different implementation approaches. The government perspective focused on comprehensive regulatory frameworks and international cooperation initiatives. The academic perspective emphasized institutional leadership and interdisciplinary research. The youth perspective prioritized individual empowerment through AI literacy and meaningful participation in governance processes. The international research perspective stressed evidence-based understanding of both benefits and risks, particularly environmental impacts and systemic effects.

## Ongoing Challenges

Several challenges were identified without complete resolution. The environmental paradox of AI – where systems designed to improve sustainability may contribute to environmental degradation through energy consumption – requires frameworks to measure net environmental impact. The crisis of AI-generated misinformation and deepfakes needs technical and policy solutions for detection and prevention at scale.

International cooperation frameworks, while widely endorsed, require detailed implementation across diverse political and economic systems. Ensuring AI literacy reaches all populations, especially in developing countries, remains a practical challenge despite being identified as essential.

## Conclusion

The discussion demonstrated understanding of AI’s complex relationship with sustainable development among key stakeholders. Rather than presenting AI simply as a sustainability tool, the conversation examined AI as a transformative force that challenges our capacity for collective action on sustainability issues.

The combination of policy, academic, youth, and international research perspectives created dialogue that addressed governance, equity, and social trust in the digital age. While significant challenges remain, the consensus on principles and complementary stakeholder perspectives suggest a foundation for continued collaboration in developing AI governance frameworks that can advance sustainable development goals while addressing risks and ensuring equitable participation in shaping our technological future.

Cynthia Su: Ladies and gentlemen, good afternoon. Welcome to this listening talk at IGF 2025, AI for Sustainable Development, the Roles of the Public and Private Sectors. I’m Cynthia Su, the Deputy Dean of Tsinghua Range Joint Research Institute for Intelligent Society at Tsinghua University, China. And I’m honored to serve as your moderator for today’s session. First of all, on behalf of the organizers, Tsinghua University, I would like to extend our warmest welcome to our distinguished guests and audience. Today’s listening talk will focus on how AI can help advance sustainable development. And today we are honored to have four distinguished guests to join us. First of all, I’d like to welcome Professor Yong Guo, the Vice Chair of University Council of Tsinghua University to deliver the opening remarks. Welcome, Professor Guo.

Yong Guo: Thank you, Xin, for your brief introduction. Distinguished guests, colleagues and friends, ladies and gentlemen, it’s a true pleasure to join you at the 20th United Nations Internet Governance Forum to explore some of the most pressing questions of our time. How can AI contribute to sustainable development? And how are the respective roles of the public and private sectors evolving in this process? On behalf of Tsinghua University, the host of this listening talk, I would like to extend my warmest welcome and heartfelt thanks to all the distinguished experts and guests for being here today. Sustainable development in the AI era has become a central theme in global governance. Chinese President Mr. Xi Jinping has emphasized that AI is a strategic technology and the forefront of today’s scientific revolution and industrial transformation, with strong spillover effects that can drive broader progress. Universities are engines for knowledge and innovation, so we must take the active role in leading AI to serve communities and accessibility. As AI is becoming more and more integrated into society, Tsinghua is committed to leading in AI research and applications. We are also pursuing interdisciplinary studies that investigate how to govern a future intelligent society. In China, we refer to this concept as intelligent society governance. We are deeply mindful of the societal challenges that AI brings. In response, Tsinghua University established the Institute for Intelligent Society Governance in 2019, with Professor Su Jun serve as its founding dean. This institute brings together the university’s multidisciplinary strategies to conduct in-depth research on the fundamental theories and core policy issues of the intelligent society. Over the years, the RISG has made significant contributions in key areas, including AI and information cocoons, social polarization, employment transformation, and energy transition. Our work represents an important exploration of Chinese solutions for intelligent society governance. These efforts have received strong support from the Chinese governments and broad recognition from across the society. The RISG has delivered exceptional outcomes in scientific research, talent cultivation, international collaboration, and standard setting, offering valuable Chinese insights to the global discourse on intelligent society governance. We are honored to host this lightning talk today, and in this context, I would like to share three reflections. First, we must focus on talent development. This means advancing interdisciplinary education in AI and intelligent society governance to cultivate a new generation of talent, individuals with both technical proficiency and a strong sense of social responsibility. Secondly, we must strengthen collaborative mechanisms. Deeper collaboration across government, industry, academia, and research institutions is needed to tackle both theoretical and practical challenges in using AI for sustainable development and to elevate our innovation and applications to new heights. Thirdly, we must deepen international cooperation. By advancing global exchanges in technology, standard setting, and financial integration, we can ensure digital dividends are shared more widely and equitably. In closing, I wish this session great success. I look forward to hearing the diverse insights from all the speakers today and to joining forces in shaping AI’s development towards a more sustainable and equitable future for all. Thank you. Thank you.

Cynthia Su: Thank you. Thank you, Professor Guo, for your inspiring remarks. Now I’d like to invite Ms. Xuanyun You, the Associate Deputy Director-General of the Bureau of Law-Based Cyberspace Governance at the Cyberspace Administration of China. She will speak about how the Chinese government is actively promoting the development of international frameworks and implementing national policies, regulations, and standards of AI towards sustainable development. Welcome.

Xuanyun You: Thank you. As you know, AI has exerted profound influence on socioeconomic development and the progress of human civilizations, bringing unprecedented development opportunities to the world. While present unprecedented risks and challenges, AI can turbocharge stable development, but transforming its potential into reality requires AI that reduces biases, misinformation, and security threats, building capacity on computing power, data, and governance, global coordinating to build a safe, secure, and inclusive AI that is accessible to all. Last year, the 17th session of the UN General Assembly adopted a resolution entitled, Seize the Opportunities of Safe, Secure, and Trustworthy AI Systems for Sustainable Development. It’s the first ever resolution negotiated at the UN General Assembly to establish a global consensus approach to AI conference. The resolution encouraged member states to promote safe, secure, and trustworthy AI systems to advance sustainable development by developing regulatory and governance approaches and frameworks related to AI systems. China is an active advocate and practitioner of global AI governance. In 2023, President Xi Jinping proposed the Global AI Governance Initiative, systematically expanding on the Chinese plan in three aspects, AI development, security, and governance. The initiative suggested that development AI should adhere to the principle of people-centered, AI for good, to ensure that AI always develops in a way that is beneficial to human civilization. We support the rule of AI in promoting sustainable development. and Tackling Global Challenges. To implement the initiative, National Technical Committee on Cyber Security and Standardization Administration of China released the AI Safety Governance Framework. In 2024, the 1978 session of the UN General Assembly unanimously adopted the resolution entitled Enhancing International Cooperation on Capacity Building of AI, proposed by China. The resolution aims to achieve inclusive, beneficial and sustainable development of AI, encourages international cooperation and practical actions to help countries, especially developing countries, strengthen their AI capacity building and support the United Nations in playing a central role in international cooperation. To implement the resolution, China proposed the AI Capacity Building Action Plan for Good and for AI, and put forward five visions and goals and attend China’s actions, which aim to bridge the AI digital device, especially to help the global South benefit actively from AI developments, and promote the implementation of the UN 2030 Agenda for Sustainable Development. China has formulated and implemented laws and regulations on the development and governance of AI. First, AI policy and rules and standards, including Next Generation AI Development Plan formulated by the State Council, six regulations and rules formulated by the Cyberspace Administration of China, such as interim measures for the management of generative AI services, measures for the identification of synthetic contacts generated by AI, and several standards, including basic security requirements for generative AI services. Second, data security governance laws and regulations related to AI, such as data security law, personal information protection law, and regulations on the management of network data security. Third, online information governance regulations and rules related to AI, such as regulation on the protection of minors in cyberspace, and the rules on the governance of online violence information. In legislation and law enforcement, China adheres to the principles of giving equal importance to development and security, innovation, and governance, takes effective measures to encourage the innovative development of AI, and implements inclusive, prudent, and classified and graded supervision of AI. In the development and application of AI, China protects personal privacy and data security, prohibits theft, alteration, leakage, and other illegal collection and use of personal information, protects international property rights, and prevents and manages risks such as false information, adverse biases, and system attacks. In the future, China will speed up the formulation and improvement of relevant policies, regulations, application specifications, and ethical standards, build a technical monitoring, risk warning, and emergency response system, improve the development and management mechanism of generative AI, establish an AI security supervision system, ensure that every stage of the AI life cycle is more safe, reliable, controllable, and equitable, and will continuously counter international exchanges and cooperation in AI development, security, and governance to promote sustainable development. Thank you so much.

Cynthia Su: Okay. Thank you, Ms. Yu. Now, I’d like to welcome Ms. Xin Yi Ding, the Youth Ambassador of the Institute of Intelligent Society, Governments of Tsinghua University, to share from the young perspective.

Xin Yi Ding: Good afternoon. Honorable guests, ladies, and gentlemen, have you ever been tricked by the information online or paused before sharing a viral news thinking it might be AI-generated? These are not just hypothetical questions, they are profound signs of a huge shift. Artificial intelligence is not just changing how we access information, but also the very concept of truth and trust. As a part of the generation that creates the future, we’re born into a coexistence with artificial intelligence. Its influence shifts every corner of society, from individuals to organizations, from governments to markets. Now, let’s be clear. AI creates opportunities for many fronts, including public service, urban governance, and sustainable development. Advanced AI models, like the one I show, the fire patterns, predicts fire patterns before it spreads. And autonomous cars, like mobile eyes, could enhance road safety by providing collision avoidance systems. And also, air-powered platforms provide educational opportunities to underserved areas. With each innovation, we make human lives more efficient, convenient, and safe. AI also creates opportunities for groups unseen by the mainstream society. In rural China, farmers started selling Mao Mao fruit, named after a golden retriever who took bites on oranges. And the story went viral online, which revitalized the economy of Zigui County in rural China. By March 2024, it generated 4 billion yuan in annual sales. Among other similar cases, AI also demonstrates the power not just to benefit a few, but also to uplift many. Yet, as we embrace AI’s promises, we must also see its risks. Data monopolies, algorithmic biases, information overload, and high energy consumption, all of which could widen the gap between the ones with power and the ones without. Let’s take this information as an example. Our generation habitually relies on AI, yet AI can generate false information while presenting it as truth. This photo, it shows me, but I’ve never had that hairstyle, never got a sweater like this. It was actually swapped with a face of the lady there, and it’s generated by Pika Art. So AI has immense power that’s misleading. Imagine seeing a pornography of a friend online. How many seconds would it take you to doubt if it’s deepfaked? When AI can so easily blur the lines between reality and falsehood, such fabrication can cause great harm. And I’m presenting a video right here. It’s IGF. It’s also AI-generated. None of this is real. And this is not just a personal, interpersonal relationship danger. It’s also a crisis in terms of public trust. When a world leader’s face can be put into any video, when seeing is no longer believing, our executional trust crumbles. And that is why our generation bears the dual responsibility to act in both public and private spheres. This begins with imperative of intelligent society governance, which is to remedy the impact of AI and putting a focus on a human and ethical aspect of AI society. And first, at individual level, we’ll be empowered with cognitive tools to guard against AI hallucination, to question sorts, cultivate critical thinking. Second, as collective, we young people cannot just settle for symbolic participations. We must also push for ethical frameworks and regulations that keep pace with AI’s rapid growth. Thirdly, we need to improve AI literacy for everyone. People need to understand how data is used and collected and learn logic behind algorithms. And our mission, as Yuan stated, it’s to leave no one behind. We have a long journey ahead in the future shaped by AI, but I believe we’re ready to take the right steps.

Cynthia Su: Thank you. Thank you, Ms. Ding, for your excellent insights on the importance of youth in promoting a sustainable development of AI. And finally, it’s my great honor to invite our last keynote speaker, the Professor Ron and Anthony Meddalia from Copenhagen Business School. Welcome.

Rony Medaglia: Thank you very much. So when we talk about artificial intelligence impacts on sustainability, of course, a useful thing to do is look at what data shows us and what scholarly research shows us. So I’ll give you a little bit of insights and you’re free to look up into the sources of everything I say in the QR codes that you see in the slides, starting from my own profile as professor at the Copenhagen Business School. When we talk about potentials, basically it’s very hard to think of any achievement of any of the SDGs without using artificial intelligence. Research shows us that the potentials are really in three areas. Efficiency, data governance and sustainable business models. So here are some examples from research. This is an example that really shows us how you can use AI for increasing efficiency for SDG goals. In the city of Aarhus in Denmark, where I live and work, there has been a use of sensors in the drinkable water grid of the city to anticipate the use of drinkable water throughout the territory and therefore reduce waste of such a precious resource. So this is clearly an example in line with SDG number six about clean water and sanitation. Another key area of use for AI is related to sustainability data governance. This is an example of a Danish company that uses drone technology to collect data about emissions from factories or from physical locations that are very hard to reach otherwise. And by feeding this data into machine learning, they’re able to anticipate levels of pollution in an area and inform policies. The third potential area is about creating new sustainable business models. This is a Danish company that is using a blockchain together with artificial intelligence to track the provenance of products such as design products, in this case a pair of shoes, whereby the consumer, by scanning a QR code, can identify whether that pair of shoes has been produced in ways that are environmentally friendly. However, the other side research shows us of AI impacts on sustainability are also negative. So we do know that, for instance, generative AI, the new large language models, use a lot of energy. So the flip side of this is that, for instance, research shows us that to generate one single image with a large language model, such as chat GPT, uses the same amount of CO2 as charging your mobile phone up to 50%. These servers that do the strong computing for large language models also need to be cooled down. So it has been calculated recently that the global AI demand may be accountable in two years from now for a water withdrawal equal to six times the whole country of Denmark for a whole year of water. Another element that cannot be understated is the so-called rebound effects. When we are able to distribute resources more efficiently, think about the sharing of cars that is done through new applications, this could be unintended consequences. For instance, people use less of public transportation, and then we will have more traffic on the streets. So how do we account for these unintended consequences? Last but not least, with AI, we see a lot of potential for bias and discrimination. If you ask a large language model to produce an image, for instance, of a CEO, you will always have, most of the times, a white young male. But if you ask for an HR manager or a nurse, you will always have a woman. So the potential for increasing discrimination, digital divide, and bias is sort of embedded in the way AI works in many cases, and it has to be mitigated. So the bottom line is that we need to have an understanding of what research shows us besides the desiderata of AI on sustainability, and therefore mitigate the risk in order to achieve the potentials of artificial intelligence for sustainable development goals. Thank you very much.

Cynthia Su: Thank you for your excellent speech. So dear guests and audience members, Today’s keynote speakers have shared from various perspectives how AI empowers sustainable development, its potential risks and practices for promoting AI sustainable growth, inspiring deep reflection among us. Due to time constraints, today’s event comes to an end. Thank you very much, and I’m looking forward to seeing you next year. The last but not the least, tomorrow we’ll have the session at 9am at Workshop 3. I’m looking forward to seeing you there as well. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you. Thank you.

Agreement points

AI has transformative potential for sustainable development across multiple sectors

Speakers

– Yong Guo
– Xuanyun You
– Xin Yi Ding
– Rony Medaglia

Arguments

AI is a strategic technology driving scientific revolution and industrial transformation with strong spillover effects

AI can turbocharge sustainable development by reducing biases, misinformation, and security threats while building capacity in computing power, data, and governance

AI creates opportunities in public service, urban governance, and sustainable development through applications like fire pattern prediction and autonomous vehicles

AI potentials for sustainability exist in three key areas: efficiency, data governance, and sustainable business models

Summary

All speakers acknowledge AI’s significant potential to advance sustainable development goals through various applications including disaster prediction, resource management, urban governance, and creating new business models

Topics

Development | Economic | Infrastructure

Multi-stakeholder collaboration is essential for effective AI governance

Speakers

– Yong Guo
– Xuanyun You
– Cynthia Su

Arguments

Deeper collaboration across government, industry, academia, and research institutions is needed to tackle AI challenges

China proposed AI Capacity Building Action Plan to bridge the digital divide and help developing countries benefit from AI developments

Academic institutions like Tsinghua University play crucial roles in hosting international dialogue on AI governance

Summary

Speakers emphasize the need for collaborative approaches involving government, industry, academia, and international institutions to address AI governance challenges effectively

Topics

Development | Legal and regulatory | Sociocultural

AI poses significant risks that must be addressed alongside its benefits

Speakers

– Xuanyun You
– Xin Yi Ding
– Rony Medaglia

Arguments

AI can turbocharge sustainable development by reducing biases, misinformation, and security threats while building capacity in computing power, data, and governance

AI presents unprecedented risks including data monopolies, algorithmic biases, information overload, and high energy consumption

AI systems exhibit embedded bias and discrimination, often reinforcing stereotypes in generated content

Summary

Speakers recognize that while AI offers tremendous opportunities, it also presents serious challenges including bias, misinformation, energy consumption, and potential for discrimination that require careful management

Topics

Human rights | Development | Cybersecurity

Similar viewpoints

Both speakers emphasize the importance of international cooperation and capacity building to ensure equitable access to AI benefits, particularly for developing countries

Speakers

– Yong Guo
– Xuanyun You

Arguments

Global exchanges in technology, standard setting, and financial integration are necessary to ensure equitable sharing of digital dividends

China proposed AI Capacity Building Action Plan to bridge the digital divide and help developing countries benefit from AI developments

Topics

Development | Economic | Infrastructure

Both speakers highlight how AI systems can perpetuate harmful biases and create misleading content, emphasizing the need for critical evaluation of AI outputs

Speakers

– Xin Yi Ding
– Rony Medaglia

Arguments

AI can generate false information and deepfakes that blur the lines between reality and falsehood, threatening public trust

AI systems exhibit embedded bias and discrimination, often reinforcing stereotypes in generated content

Topics

Human rights | Sociocultural

Both speakers emphasize the critical role of universities and academic institutions in leading AI research, governance discussions, and ensuring AI serves broader societal needs

Speakers

– Yong Guo
– Cynthia Su

Arguments

Universities must take active roles in leading AI to serve communities and accessibility through interdisciplinary research

Academic institutions like Tsinghua University play crucial roles in hosting international dialogue on AI governance

Topics

Development | Sociocultural

Unexpected consensus

Environmental impact of AI systems

Speakers

– Xin Yi Ding
– Rony Medaglia

Arguments

AI presents unprecedented risks including data monopolies, algorithmic biases, information overload, and high energy consumption

Generative AI uses significant energy resources and creates substantial water withdrawal demands for server cooling

Explanation

Despite representing different perspectives (youth advocate vs. academic researcher), both speakers independently identified AI’s environmental impact as a significant concern, with specific attention to energy consumption – an issue that might not be immediately obvious when discussing AI for sustainability

Topics

Development | Infrastructure

Need for comprehensive regulatory frameworks

Speakers

– Xuanyun You
– Xin Yi Ding

Arguments

China has formulated comprehensive laws and regulations including AI policy rules, data security governance, and online information governance

Young people bear dual responsibility to act in both public and private spheres regarding AI governance

Explanation

Unexpected alignment between a government official emphasizing existing regulatory frameworks and a youth representative calling for stronger ethical frameworks and regulations, showing cross-generational agreement on the need for robust governance structures

Topics

Legal and regulatory | Human rights

Overall assessment

Summary

The speakers demonstrated strong consensus on AI’s transformative potential for sustainable development, the need for multi-stakeholder collaboration, and the importance of addressing AI risks alongside benefits. Key areas of agreement include the necessity of international cooperation, the role of academic institutions, and the recognition of both opportunities and challenges presented by AI systems.

Consensus level

High level of consensus with complementary perspectives rather than conflicting viewpoints. The speakers approached the topic from different angles (government policy, academic research, youth advocacy, international cooperation) but arrived at similar conclusions about the need for balanced, collaborative approaches to AI governance for sustainable development. This strong alignment suggests a mature understanding of AI governance challenges and indicates potential for effective policy coordination across different stakeholder groups.

Different viewpoints

Approach to AI risk mitigation and regulation

Speakers

– Xuanyun You
– Rony Medaglia

Arguments

China has formulated comprehensive laws and regulations including AI policy rules, data security governance, and online information governance

AI systems exhibit embedded bias and discrimination, often reinforcing stereotypes in generated content

Summary

You emphasizes China’s comprehensive regulatory framework as a solution to AI challenges, while Medaglia focuses on the inherent technical limitations and biases embedded in AI systems that require mitigation beyond regulatory approaches

Topics

Legal and regulatory | Human rights

Primary focus for addressing AI challenges

Speakers

– Xin Yi Ding
– Yong Guo

Arguments

The younger generation needs cognitive tools to guard against AI hallucination and must cultivate critical thinking

Universities must take active roles in leading AI to serve communities and accessibility through interdisciplinary research

Summary

Ding emphasizes individual-level cognitive skills and critical thinking as primary solutions, while Guo focuses on institutional leadership and interdisciplinary research approaches

Topics

Sociocultural | Development

Unexpected differences

Environmental impact prioritization

Speakers

– Rony Medaglia
– Other speakers

Arguments

Generative AI uses significant energy resources and creates substantial water withdrawal demands for server cooling

Explanation

Medaglia was the only speaker to explicitly quantify and emphasize the environmental costs of AI operations, while other speakers focused primarily on AI’s potential benefits for sustainability without addressing its environmental footprint. This represents an unexpected gap in a discussion about sustainable development

Topics

Development | Infrastructure

Overall assessment

Summary

The discussion showed relatively low levels of direct disagreement, with speakers generally complementing rather than contradicting each other’s perspectives. Main areas of difference centered on implementation approaches (regulatory vs. technical vs. educational) and priority focus (institutional vs. individual vs. international cooperation)

Disagreement level

Low to moderate disagreement level. The speakers represented different stakeholder perspectives (government, academia, youth, international research) and their differences were more about emphasis and approach rather than fundamental opposition. This suggests a constructive dialogue environment but may indicate insufficient critical examination of competing approaches to AI governance and sustainable development

Partial agreements

Similar viewpoints

Both speakers emphasize the importance of international cooperation and capacity building to ensure equitable access to AI benefits, particularly for developing countries

Speakers

– Yong Guo
– Xuanyun You

Arguments

Global exchanges in technology, standard setting, and financial integration are necessary to ensure equitable sharing of digital dividends

China proposed AI Capacity Building Action Plan to bridge the digital divide and help developing countries benefit from AI developments

Topics

Development | Economic | Infrastructure

Both speakers highlight how AI systems can perpetuate harmful biases and create misleading content, emphasizing the need for critical evaluation of AI outputs

Speakers

– Xin Yi Ding
– Rony Medaglia

Arguments

AI can generate false information and deepfakes that blur the lines between reality and falsehood, threatening public trust

AI systems exhibit embedded bias and discrimination, often reinforcing stereotypes in generated content

Topics

Human rights | Sociocultural

Both speakers emphasize the critical role of universities and academic institutions in leading AI research, governance discussions, and ensuring AI serves broader societal needs

Speakers

– Yong Guo
– Cynthia Su

Arguments

Universities must take active roles in leading AI to serve communities and accessibility through interdisciplinary research

Academic institutions like Tsinghua University play crucial roles in hosting international dialogue on AI governance

Topics

Development | Sociocultural

Key takeaways

AI has dual potential for sustainable development – offering significant opportunities through efficiency gains, improved data governance, and new business models, while also presenting serious risks including energy consumption, bias, and misinformation

Successful AI governance requires a multi-stakeholder approach involving government, industry, academia, and civil society working together across national boundaries

China is positioning itself as a leader in AI governance through comprehensive policy frameworks, international initiatives, and the concept of ‘intelligent society governance’

Youth perspectives are critical for AI governance as the generation most affected by AI integration, emphasizing the need for AI literacy, critical thinking, and ethical frameworks

International cooperation and capacity building are essential to ensure AI benefits are shared equitably, particularly helping developing countries bridge the digital divide

Real-world applications demonstrate AI’s tangible benefits for sustainability, from water management in Danish cities to economic revitalization in rural China

The challenge lies in maximizing AI’s positive potential while effectively mitigating risks through proper governance, regulation, and education

Resolutions and action items

China will speed up formulation and improvement of AI policies, regulations, application specifications, and ethical standards

Build technical monitoring, risk warning, and emergency response systems for AI

Improve development and management mechanisms for generative AI and establish AI security supervision systems

Continue international exchanges and cooperation in AI development, security, and governance

Advance interdisciplinary education in AI and intelligent society governance to cultivate talent with technical proficiency and social responsibility

Implement the AI Capacity Building Action Plan to help developing countries strengthen AI capabilities

Promote AI literacy improvement for everyone to understand data usage and algorithmic logic

Unresolved issues

How to effectively address the rebound effects and unintended consequences of AI efficiency gains

Specific mechanisms for ensuring AI systems remain ‘safe, reliable, controllable, and equitable’ throughout their lifecycle

Concrete strategies for mitigating AI’s significant energy consumption and environmental impact

Detailed frameworks for international coordination on AI standards and governance across different political and economic systems

Methods for effectively combating deepfakes and AI-generated misinformation while preserving innovation

Specific approaches for eliminating algorithmic bias and discrimination embedded in AI systems

How to balance AI innovation with prudent regulation without stifling technological progress

Suggested compromises

China’s approach of ‘giving equal importance to development and security, innovation and governance’ as a balanced regulatory philosophy

Implementation of ‘inclusive, prudent, and classified and graded supervision of AI’ rather than blanket restrictions

Focus on capacity building and international cooperation to address the digital divide rather than restricting AI development

Emphasis on education and AI literacy as a complement to regulatory approaches

Multi-stakeholder governance involving both public and private sectors rather than government-only or market-only solutions

Have you ever been tricked by the information online or paused before sharing a viral news thinking it might be AI-generated? These are not just hypothetical questions, they are profound signs of a huge shift. Artificial intelligence is not just changing how we access information, but also the very concept of truth and trust.

Speaker

Xin Yi Ding

Reason

This opening immediately reframes the discussion from abstract policy considerations to visceral, personal experiences that resonate with everyone. It introduces the fundamental epistemological crisis that AI creates – challenging our basic ability to distinguish truth from falsehood, which is foundational to all other discussions about AI governance and sustainability.

Impact

This comment shifted the discussion from high-level policy frameworks to concrete, relatable concerns. It established a more urgent tone and introduced the critical theme of trust erosion that underlies many AI governance challenges, setting up the need for the practical solutions discussed later.

When AI can so easily blur the lines between reality and falsehood, such fabrication can cause great harm… When a world leader’s face can be put into any video, when seeing is no longer believing, our executional trust crumbles.

Speaker

Xin Yi Ding

Reason

This insight connects individual-level deception (deepfakes of friends) to systemic threats to democratic governance and public trust. It demonstrates how AI’s impact on sustainability isn’t just about energy consumption or efficiency, but about the fundamental social fabric that enables collective action on sustainability challenges.

Impact

This comment elevated the discussion beyond technical solutions to examine the societal prerequisites for sustainable development. It highlighted how AI threatens the social trust necessary for collective action on sustainability, adding a crucial dimension that the other speakers hadn’t addressed.

However, the other side research shows us of AI impacts on sustainability are also negative… to generate one single image with a large language model uses the same amount of CO2 as charging your mobile phone up to 50%

Speaker

Rony Medaglia

Reason

This provides concrete, quantifiable data that starkly illustrates AI’s environmental costs in terms everyone can understand. It transforms abstract concerns about energy consumption into tangible comparisons, making the sustainability paradox of AI viscerally clear.

Impact

This comment introduced hard data that balanced the optimistic framing from earlier speakers. It forced a reckoning with AI’s direct environmental costs and established the need for the nuanced approach to AI governance that considers both benefits and costs simultaneously.

Another element that cannot be understated is the so-called rebound effects. When we are able to distribute resources more efficiently… this could be unintended consequences. For instance, people use less of public transportation, and then we will have more traffic on the streets.

Speaker

Rony Medaglia

Reason

This introduces the sophisticated concept of rebound effects – how efficiency gains can paradoxically lead to increased resource consumption. It challenges the linear thinking that efficiency automatically equals sustainability and introduces systems thinking to the discussion.

Impact

This comment added crucial complexity to the discussion by showing how AI’s benefits can backfire in unexpected ways. It demonstrated the need for holistic, systems-level thinking in AI governance rather than focusing solely on direct effects, influencing how we must approach AI policy for sustainability.

AI has immense power that’s misleading… our generation bears the dual responsibility to act in both public and private spheres… We young people cannot just settle for symbolic participations. We must also push for ethical frameworks and regulations that keep pace with AI’s rapid growth.

Speaker

Xin Yi Ding

Reason

This rejects tokenistic youth involvement and demands substantive participation in AI governance. It connects generational responsibility with the urgency of AI development, arguing that those who will live longest with AI’s consequences must have real agency in shaping its development.

Impact

This comment challenged traditional power structures in technology governance and established youth not as beneficiaries of adult decision-making, but as essential actors with unique stakes and perspectives. It added urgency to the governance discussion by emphasizing the pace of AI development relative to regulatory responses.

Overall assessment

These key comments fundamentally transformed what could have been a conventional policy discussion into a nuanced examination of AI’s paradoxical relationship with sustainability. Xin Yi Ding’s contributions introduced existential questions about truth and trust that reframed sustainability as not just an environmental challenge but a social and epistemological one. Rony Medaglia’s data-driven insights provided the empirical grounding that balanced optimistic policy statements with hard realities about AI’s environmental costs and systemic complexities. Together, these comments created a more sophisticated understanding of AI governance that moves beyond simple benefit-risk calculations to consider feedback loops, unintended consequences, and the social prerequisites for sustainable development. The discussion evolved from presenting AI as a tool for sustainability to examining AI as a force that fundamentally challenges our capacity for collective action on sustainability challenges.

How can we effectively measure and mitigate the unintended consequences of AI implementation, particularly rebound effects in sustainability initiatives?

Speaker

Rony Medaglia

Explanation

Professor Medaglia highlighted rebound effects as a significant concern, using the example of car-sharing apps potentially reducing public transportation use and increasing traffic, but didn’t provide solutions for addressing these unintended consequences.

What specific mechanisms and frameworks are needed to ensure AI literacy education reaches all populations, especially in developing countries?

Speaker

Xin Yi Ding

Explanation

Ms. Ding emphasized the need for AI literacy for everyone but didn’t elaborate on the practical implementation strategies or how to overcome barriers to access, particularly in underserved communities.

How can we quantify and balance the environmental costs of AI (energy consumption, water usage) against its sustainability benefits?

Speaker

Rony Medaglia

Explanation

Professor Medaglia presented concerning statistics about AI’s environmental impact (CO2 emissions, water usage) but didn’t address how to create frameworks for measuring net environmental impact when AI is used for sustainability purposes.

What are the most effective methods for detecting and preventing AI-generated misinformation and deepfakes at scale?

Speaker

Xin Yi Ding

Explanation

Ms. Ding demonstrated the ease of creating convincing deepfakes and highlighted the crisis of public trust, but didn’t discuss technical or policy solutions for detection and prevention of such content.

How can international cooperation frameworks be strengthened to ensure equitable AI development and governance across different political and economic systems?

Speaker

Yong Guo and Xuanyun You

Explanation

Both speakers emphasized the importance of international cooperation and China’s initiatives, but didn’t address the practical challenges of implementing global governance frameworks across diverse political systems and economic development levels.

What specific metrics and evaluation methods should be used to assess the effectiveness of AI capacity building initiatives in developing countries?

Speaker

Xuanyun You

Explanation

Ms. You mentioned China’s AI Capacity Building Action Plan and efforts to help the Global South, but didn’t specify how success would be measured or what indicators would demonstrate effective capacity building.