Access Partnership participated in the AI for Good Global Summit 2024, held on 29-31 May in Geneva, which brought together over 10,000 in-person and 25,000 online participants from 145 countries. Organised by the International Telecommunication Union (ITU) in partnership with 40 United Nations (UN) sister agencies, the annual summit focused on building a global framework for AI and identifying practical AI applications to accelerate progress towards the UN Sustainable Development Goals (SDGs).
Global AI governance
The summit underscored the crucial role of international collaboration in establishing a comprehensive global framework for AI governance. All countries supported greater inclusivity in AI development and governance, with some advocating for flexibility based on local contexts while others emphasised the need for regulatory convergence. Global AI policy leadership activities were also showcased, from the Bletchley Declaration and the Hiroshima AI process to the GPAI Delhi Summit.
Transparency and standards: Building trust
The summit emphasised the importance of transparency and robust standards in AI development. Discussions explored challenges in model interpretability, particularly for complex AI models, while highlighting ongoing efforts to establish international standards for areas like AI trustworthiness, bias mitigation, and data quality.
AI regulation
Experts called for regulations to cover the wider AI ecosystem, including user behaviour, not just models. They advocated for increased spending on AI safety and discussed the allocation of regulatory obligations within the AI value chain. The World Intellectual Property Office (WIPO) advocated for a human-centric IP regime.
AI risks
Key risks include mis/disinformation, deepfakes, and scams. The International Labour Organization also highlighted that AI would create more jobs, but some experts expressed concerns about the distribution of jobs and profits. Mitigating risks of AI use in the financial sector, through stronger social safety nets and upskilling, was also emphasised.
Striking a balance: Innovation and risk mitigation
A prominent theme throughout the summit was the need to balance innovation with risk mitigation. Discussions around AI regulation emphasised the importance of fostering a regulatory environment that encourages responsible AI development while ensuring potential risks are addressed. This includes concerns surrounding bias in AI models, the ethical use of data, and the potential for job displacement.
AI models
Transparency and model evaluation frameworks were deemed critical. Testing and standards development, particularly for AI trustworthiness, interpretability, bias treatment, sustainability, human oversight, functional safety, and data quality, are considered crucial areas. China emphasised a multi-stakeholder, science-based approach, while Japan supported voluntary standards. The International Organization for Standardization called for coherence across standard development efforts. Finally, some experts cautioned against the over-regulation of open-source AI.
AI for a sustainable future
AI’s potential to accelerate progress towards the SDGs was a major focus. From improving healthcare access to addressing climate change challenges, AI offers a powerful toolkit for tackling global issues. The summit witnessed various demos showcasing how AI is used for a positive social impact.
The road ahead
The AI for Good Global Summit 2024 served as a catalyst for discussions on how to navigate the evolving landscape of AI governance. The Summit provided valuable insights on the need for international collaboration on a global AI framework, balancing innovation with risk mitigation, and prioritising transparency and standards for responsible AI adoption – with the greater goal of allowing AI to “turbocharge sustainable development”, in the words of UN Secretary-General Antonio Guterres.
Access Partnership is working closely with governments and the private sector to shape AI policy and regulation globally. To learn more about staying ahead of AI policy and regulatory developments, please contact Bensen Koh at [email protected] or Dana Ramadan at [email protected].
