data jungle adventures

The AI for Good Global Summit is the leading United Nations platform for inclusive dialogue and collaborative action on artificial intelligence. Bringing together innovators, policymakers, industry leaders, and civil society, the summit aims to ensure that AI technologies accelerate progress toward the UN Sustainable Development Goals.

This year’s 2025 summit, held from July 8 to 11, features dozens of high-level panels, talks, and workshops across disciplines, targeting, among other subjects, climate and ethics.

On its third day, the spotlight was on a topic growing increasingly urgent: AI governance. With the pace of AI development outstripping policy responses, this day’s sessions focused on how we can collectively build frameworks that are both future-proof and globally inclusive. Here is a glimpse of three sessions.

1. Laying the Foundations for AI Governance

This panel brought together key global figures, including George Papandreou, Lan Xue, Robert Trager, Artemis Seaford, and Dawn Song. One of the key topics discussed on this panel was the obstacles hindering the move from Principles to Policy and Practice in AI Governance. The list included the following:

• The challenge of translating principles into practice, particularly across borders and sectors.
• Geopolitical barriers to cooperation and the risks of nationalistic AI development.
• Concerns about the centralisation of power, especially where financial decisions influence regulatory direction.
• The mismatch between policy timelines and technological acceleration, highlighting the need for science-informed governance.

2. What Frontier AI Means for AI Governance

This focused talk, delivered by Daniela Rus from the Computer Science and Artificial Intelligence Laboratory (CSAIL) of MIT, provided a deep dive into the implications of cutting-edge AI developments.

Frontier AI systems—typically developed through pretraining and finetuning—are expanding the limits of what AI can achieve. Foundation models demonstrate remarkable capabilities in speed, insight, creativity, foresight, and even empathy, making them truly transformative. Yet, their advancement faces challenges such as data quality, computational intensity, and interpretability. Drawing inspiration from nature offers promising solutions: for example, the C. elegans worm has inspired the creation of Liquid Neural Networks—models that are not only small and energy efficient but also capable of dynamic reasoning. Supporting such innovation is key to shaping a more intelligent and sustainable AI future.

3. How to Make AI Governance Fit for Purpose

In this dynamic panel, participants included Zhongde Shan, Jennifer Bachus, Gabriela Ramos, Chuen Hong Lew, and Anne Bouverot.
A key message highlighted by Jennifer Bachus was the importance of multi-stakeholder approaches in shaping AI governance, ensuring that voices from government, industry, academia, and civil society are represented. She emphasised that AI should be harnessed to improve the well-being of people and nations, not just to drive efficiency or profit. At the same time, there was a strong caution against excessive regulation, which could hinder innovation and stall the progress of what remains a transformative and rapidly evolving industry.

Final reflections

On Day 3 of the AI for Good conference, focused on AI for Governance, a key message was the importance of responsible and transparent AI use. Just as clear communication relies on well-curated words, images, and clear messages, governance must guide AI development ethically. Similar to how regulators promote recycling and public transport to reduce environmental impact, policymakers must steer AI toward fairness, sustainability, and societal benefit.

As the summit made clear, AI governance is no longer optional—it is essential. While there is growing consensus on the need for guardrails, there are also valid concerns about over-regulation dampening innovation.

Inclusion, a recurring theme, reminds us that AI must be shaped by diverse perspectives. And though the challenges ahead are formidable, from geopolitics to enforcement, the summit left participants with a sense of collective responsibility—and the tools to begin moving forward.

Feature image: Midjourney