Session
Organizer 1: Technical Community, Western European and Others Group (WEOG)
Organizer 2: Civil Society, African Group
Organizer 3: Private Sector, Latin American and Caribbean Group (GRULAC)
Organizer 4: Technical Community, Western European and Others Group (WEOG)
Organizer 5: Civil Society, Asia-Pacific Group
Organizer 6: Civil Society, Asia-Pacific Group
Speaker 1: Ke GONG, Technical Community, Western European and Others Group (WEOG)
Speaker 2: Daisy Selematsela, Civil Society, African Group
Speaker 3: Ricardo Israel Robles Pelayo, Private Sector, Latin American and Caribbean Group (GRULAC)
Speaker 4: Horst Kremers, Technical Community, Western European and Others Group (WEOG)
Speaker 5: Xiaofeng Tao, Civil Society, Asia-Pacific Group
Roundtable
Duration (minutes): 60
Format description: The roundtable format promotes equal - footing conversations, empowering a wide range of stakeholders to voice their opinions without hierarchy. Its circular arrangement not only heightens interaction and teamwork but also paves the way for strong connections among participants. This setup is ideal for delving deep into complex issues, driving higher levels of engagement. Moreover, with participants from diverse backgrounds, it sparks a rich exchange of ideas, often resulting in innovative solutions. Spanning 90 minutes, the event is structured to balance thoroughness and flexibility. In the first 30 minutes, speakers present best practices in detail. The next 40 minutes are dedicated to lively discussions, interactive Q&A sessions moderated by the host, and breakout groups involving both onsite and online audiences. The final 20 minutes are for distilling actionable results. This time - management approach ensures that all aspects of the topic are explored, keeping the hybrid audience engaged throughout.
1. Inclusive and responsible development: What policies can prevent AI innovations from exacerbating digital divides, particularly in the Global South? 2.Governance: Given the cross-border nature of AI development, how can national and international governance frameworks, in line with the Global Digital Compact, ensure ethical and responsible AI innovation? 3. Stakeholder Engagement: What are the most effective ways to engage diverse stakeholders in the decision-making process of AI development to ensure ethical and responsible outcomes without stifling innovation?
What will participants gain from attending this session? Participants will gain a deeper understanding of aligning AI innovation with the GDC and UNESCO’s ethical guidelines and learn global best practices for ethical and responsible AI innovations. Through sharing sessions, they can exchange experiences and engage in cross-sector dialogue with international peers to get insights into effective approaches to develop AI governance models that balance innovation with accountability, and also identify opportunities for global partnerships. Additionally, they will be able to develop strategies for integrating ethical considerations into their own AI-related projects, ultimately contributing to a more responsible and ethical AI ecosystem. They also get the opportunity to contribute to actionable recommendations and co-create strategies for harmonizing AI policies with SDGs.
1. No Poverty
1.4
1.b
4. Quality Education
4.4
4.7
5. Gender Equality
5.b
5.c
9. Industry, Innovation and Infrastructure
9.4
9.5
9.b
9.c
10. Reduced Inequalities
10.2
10.3
11. Sustainable Cities and Communities
11.7
11.a
11.c
16. Peace, Justice and Strong Institutions
16.10
17. Partnerships for the Goals
Description:
The rapid advancement of Artificial Intelligence (AI) has catalyzed transformative innovations, which is reshaping multiple aspects of our lives. From technological innovation of new Algorithms to advancements in AI hardware, from Generative AI that sparks innovative content creation to open-source platforms enabling collaborative development of transparent systems, AI innovations are increasingly deployed in critical sectors such as healthcare, climate resilience, and education, driving economic growth, enhancing social well – being. AI innovation not only provides breakthrough solutions to address complex global challenges but also demonstrating immense potential to accelerate progress toward the Sustainable Development Goals (SDGs). However, AI innovations also introduce profound ethical, responsibility and governance challenges. Algorithmic biases risk perpetuating systemic inequalities, while data privacy breaches threaten individual rights. Moreover, the digital divide risks widening as marginalized communities struggle to access AI tools, undermining the inclusive vision of the Global Digital Compact (GDC). The GDC emphasizes the need for inclusive multistakeholder cooperation to ensure digital technologies, including AI, are harnessed for the common good, and advocates for the responsible development of AI. UNESCO's Recommendation on the Ethics of AI provides a crucial framework for ethical AI development, focusing on transparency, fairness, and human-centric values. Building on the GDC's call for inclusive digital governance and UNESCO’s guidelines, this workshop aims to address these tensions by convening stakeholders to align AI innovation with responsible development and ethical imperatives. Best practices globally in responsible AI development will be showcased to highlight the diverse innovations in AI and their real-world implications, in-depth discussions will be facilitated to chart a path where AI innovation not only drives progress but also upholds the principles of ethics and responsibility. Key topics include fostering multi-stakeholder collaboration to design ethical frameworks, leveraging AI for social good, and ensuring equitable access to AI advancements.
1.Best Practices Compilation: Gather and publish best practices in responsible AI innovation, sourced from workshop participants, to promote knowledge sharing. 2.Actionable Guidelines: Develop practical guidelines for ethical AI development, taking into account the requirements of UNESCO's recommendation and the Global Digital Compact. 3.Network Building: Foster a network of stakeholders committed to ethical AI innovation, enabling continued collaboration and knowledge exchange. 4. Policy Roadmap: Propose policy recommendations to to enhance ethical and responsible AI development at national and international levels and harmonize national AI strategies with the GDC and UNESCO principles.
Hybrid Format: Under the moderator's guidance, both online and onsite speakers and attendees can speak, discuss, and ask questions. There'll be a Q&A and discussion session for everyone to share views. An online moderator will manage questions from online attendees shown on - screen. Zoom will mix onsite and virtual participants. After the session, recordings and summaries will be shared on the IGF platform and social media.
Report
Key Takeaways
1. Multifaceted Ethical Imperatives in AI Development
AI advancement extends beyond technical innovation, intersecting deeply with ethical challenges. Data biases in training models lead to unfair outcomes, while algorithmic opacity in deep learning raises accountability issues, particularly in critical sectors like healthcare and finance. Additionally, malicious uses highlight the need to embed ethics throughout the AI lifecycle—from data collection to deployment.
2. Urgency of Global Collaborative Governance
AI's borderless nature necessitates cross-national cooperation, as no single country can address its complexities alone. Developed nations often prioritize innovation speed, while developing countries focus on equitable access and risk mitigation. International frameworks must set common standards for AI safety, security, and ethics, integrating diverse stakeholder inputs.
3. Balancing Innovation with Responsible Development
While AI drives progress, irresponsible development risks job displacement, privacy violations, and environmental harm. Policymakers must incentivize innovation while enforcing regulations to ensure ethical compliance, fostering a landscape where AI's benefits are maximized without compromising societal well-being.
Call to Action
1. Strengthen Cross-Sector Ethical Review Mechanisms
Research institutions, tech companies, and governments should establish interdisciplinary ethics boards—including ethicists, technologists, and community representatives—to review AI projects at all stages. For example, financial AI services must undergo pre-deployment audits for data bias and algorithmic transparency, with public reporting to build trust.
2. Accelerate Capacity Building in Developing Regions
Developed nations and international organizations should collaborate to bridge the AI resource gap through initiatives like funding research facilities, offering scholarships, and organizing training workshops. This fosters an inclusive AI ecosystem, integrating diverse perspectives to drive global innovation.
3. Advocate for Holistic International Governance Frameworks
Governments must lead efforts within forums like the IGF to craft global AI governance frameworks covering safety, data protection, and ethical guidelines . Civil society and the private sector should contribute practical insights, ensuring frameworks adapt to rapid technological advancements.
Context and Framing
Co-organized by CAST UN Consultative Committee on Information Technology (CCIT) and Internet Society of China (ISC), this workshop convened global experts to address AI’s ethical challenges. Moderator Mr. Dai Wei (Deputy Secretary-General of Internet Society of China & Deputy Director of China IGF) opened by stressing AI’s "transformative power" and the non-negotiable need for "ethical imperatives and global inclusivity."
Prof. Gong Ke (Chair of CCIT) framed AI’s dual nature: while revolutionizing sectors from healthcare to agriculture, its opaque algorithms risk deepening global inequities. He advocated tripartite action: (1) Policy safeguards for equitable technology access, (2) Harmonization with UNESCO’s ethics framework, and (3) Cross-border stakeholder collaboration to transcend geopolitical divides.
Mr. Huang Chengqing (Vice President of Internet Society of China & Director of China IGF) positioned AI as a tool for human welfare, citing China’s Global AI Governance Initiative and emphasizing industry self-regulation. He cautioned against "innovation without ethical scaffolding," urging balanced development that prioritizes social well-being over unchecked technological advancement.
Expert Perspectives
- Guilherme Canela De Souza Godoi (Director for Digital Inclusion, Policy and Transformation & IFAP Secretary, UNESCO)
UNESCO’s mandate bridges innovation and human rights, asserting that ethical AI and technological progress are synergistic rather than contradictory. The organization’s 2021 Ethics Recommendation – ratified by 194 states – provides a human-rights-based framework grounded in 80 years of tech-ethics evolution. Current priorities include implementing Readiness Assessment Methodology in 30+ countries to evaluate governance capabilities, establishing training programs for civil servants, and addressing exclusion risks for linguistically/physically marginalized groups. Crucially, Godoi emphasized that "innovation failing to serve all communities is fundamentally illegitimate," advocating for multilingual AI systems and disability-inclusive design. - Prof. Yik Chan Chin (Associate Professor, Beijing Normal University)
Generative AI poses acute risks in education: algorithmic black boxes propagate unreliable content, intellectual property violations undermine academic integrity, and uneven access accelerates "digital poverty" between nations. Comparative analysis revealed divergent approaches. To counter these challenges, Chin urged multistakeholder dialogues on long-term cognitive impacts, strict age-gating for AI tools, and data ownership reforms preventing corporate monopolization of educational resources. - Prof. Ricardo Israel Robles Pelayo (Professor of Escuela Bancaria y Comercial, Mexico)
Mexico’s hasty AI integration exemplifies systemic risks: in education, tools exacerbate student anxiety and erode critical thinking; in justice, algorithms trained on biased data automate discrimination within fragile judicial systems. Pelayo critiqued the trend of "delegating human judgment to machines". His solution centers on legally mandating critical thinking as an ethical mediator – requiring human oversight to contextualize AI outputs through moral deliberation and rights-based assessment. - Dr. Daisy Selematsela (University of the Witwatersrand: Library)
Academic libraries face dual pressures: AI promises enhanced accessibility (personalized research support, speech-to-text tools for disabled users) but threatens job displacement, metadata interoperability, and collection security. Budget constraints heighten these tensions, particularly in Global South institutions. Selematsela proposed a balanced framework: leveraging AI for repository management (automated cataloging, content summarization) while implementing strict encryption protocols and algorithmic transparency measures. She stressed that "libraries must remain human-curated spaces," with AI augmenting – not replacing – librarians’ roles in knowledge stewardship. - Ms. Zhang Xiao (Vice President of China Internet Network Information Center (CNNlC), member of the lGF Multistakeholder Advisory Group (MAG), Executive Deputy Director of China lGF)
Governance and innovation are interdependent forces – where ethical guardrails enable responsible advancement. Drawing lessons from internet governance, Zhang advocated adapting multi-stakeholder models to AI’s "vertical" societal penetration. Her proposal centered on three pillars: (1) Developing shared vocabularies around human-centric metrics, (2) Prioritizing capacity-building to foster public awareness of AI rights, and (3) Creating interoperable standards that respect cultural diversity while preventing harmful applications.Synthesis and Conclusions
The dialogue converged on three imperatives:
Inclusion as Cornerstone: Prevent AI from exacerbating educational/resource disparities through equitable access policies.
Global Governance Alignment: Anchor national regulations in UNESCO’s framework while allowing regional adaptations (e.g., China’s pedagogical guidelines).
Ethical Innovation Protocols: Embed human oversight in high-risk domains – especially justice and education.
