At the Internet Governance Forum (IGF) 2025 in Norway, an expert panel convened to examine the growing complexity of artificial intelligence governance. The discussion, moderated by Kathleen Ziemann from the German development agency GIZ and Guilherme Canela of UNESCO, featured a rich exchange between government officials, private sector leaders, civil society voices, and multilateral organisations.
The session highlighted how AI governance is becoming a crowded yet fragmented space, shaped by overlapping frameworks such as the OECD AI Principles, the EU AI Act, UNESCO’s recommendations on AI ethics, and various national and regional strategies. While these efforts reflect progress, they also pose challenges in terms of coordination, coherence, and inclusivity.
Melinda Claybaugh, Director of Privacy Policy at Meta, noted the abundance of governance initiatives but warned of disagreements over how AI risks should be measured. ‘We’re at an inflection point,’ she said, calling for more balanced conversations that include not just safety concerns but also the benefits and opportunities AI brings. She argued for transparency in risk assessments and suggested that existing regulatory structures could be adapted to new technologies rather than replaced.
In response, Jhalak Kakkar, Executive Director at India’s Centre for Communication Governance, urged caution against what she termed a ‘false dichotomy’ between innovation and regulation. ‘We need to start building governance from the beginning, not after harms appear,’ she stressed, calling for socio-technical impact assessments and meaningful civil society participation. Kakkar advocated for multi-stakeholder governance that moves beyond formality to real influence.
Mlindi Mashologu, Deputy Director-General at South Africa’s Ministry of Communications and Digital Technology, highlighted the importance of context-aware regulation. ‘There is no one-size-fits-all when it comes to AI,’ he said. Mashologu outlined South Africa’s efforts through its G20 presidency to reduce AI-driven inequality via a new policy toolkit, stressing human rights, data justice, and environmental sustainability as core principles. He also called for capacity-building to enable the Global South to shape its own AI future.
Jovan Kurbalija, Executive Director of the Diplo Foundation, brought a philosophical lens to the discussion, questioning the dominance of ‘data’ in governance frameworks. ‘AI is fundamentally about knowledge, not just data,’ he argued. Kurbalija warned against the monopolisation of human knowledge and advocated for stronger safeguards to ensure fair attribution and decentralisation.
The need for transparency, explainability, and inclusive governance remained central themes. Participants explored whether traditional laws—on privacy, competition, and intellectual property—are sufficient or whether new instruments are needed to address AI’s novel challenges.
Audience members added urgency to the discussion. Anna from Mexican digital rights group R3D raised concerns about AI’s environmental toll and extractive infrastructure practices in the Global South. Pilar Rodriguez, youth coordinator for the IGF in Spain, questioned how AI governance could avoid fragmentation while still respecting regional sovereignty.
The session concluded with a call for common-sense, human-centric AI governance. ‘Let’s demystify AI—but still enjoy its magic,’ said Kurbalija, reflecting the spirit of hopeful realism that permeated the discussion. Panelists agreed that while many AI risks remain unclear, global collaboration rooted in human rights, transparency, and local empowerment offers the most promising path forward.
Track all key moments from the Internet Governance Forum 2025 on our dedicated IGF page.
