Values Exchange

Countries grapple with shaping AI governance amid heightened global tensions

26 May 2024
Senior Thinker (2758 XP)
Please login to save to your library
Countries grapple with shaping AI governance amid heightened global tensions

17 May 2024, North Rhine-Westphalia, Cologne: A person works at a computer with an illustrative image generated by artificial intelligence on the screen, showing code from various programming languages and a neural network diagram. (Keystone/DPA/Oliver Berg)




Countries grapple with shaping AI governance amid heightened global tensions

By Michelle Langrand

After twenty years of unsuccessful attempts at governing the internet, similar discussions over artificial intelligence are gathering pace. While the challenges may be new, old divisions are resurfacing.

As AI develops at a fast pace and seeps into more aspects of our daily lives, conversations about whether to regulate it and, how to do so are heating up. In less than four months, UN member states are set to adopt a political declaration meant to lay the tracks for the future of global governance of digital technologies and artificial intelligence. This week in Geneva, it will be the burning question at the World Society Information Summit (WSIS) Forum.

But amid geopolitical tensions at historical levels and leading powers butting heads as they engage in an AI arms race, the prospects of any major political convergence are meagre. For observers who have become accustomed to the complex and entangled web that is digital governance, it may not be such a bad thing.

20 years of talking

This year marks 20 years of the first landmark conference which bequeathed the forum its name – WSIS. In 2003, governments came together to try to decide how the world should govern its latest revolutionary innovation that would later upend all aspects of society – the internet. Divergences between those who thought companies should be left to run free and those who thought the UN should come up with some global rules to level the playing field could only be resolved by compromise. Enter the WSIS forum and the Internet Governance Forum, a “carefully crafted compromise” which for the last two decades has served as a place for dialogue and exchange of ideas in the absence of hard rules.

Read more: UN touts initiative to regulate AI at tech summit in Geneva

No big decisions or announcements should be expected at the end of this week from the WSIS forum, But to the discerning eye, plenty can be revealed. “This is the last major gathering before the Summit of the Future, and therefore countries will make their positions known, signal what they want and what the red lines are,” Jovan Kurbalija, executive director of DiploFoundation, told Geneva Solutions.

Read more: The uncertain prospects of the UN’s future summit

Geopolitical tensions

After mulling over how to regulate the digital realm for two decades, the world finds itself at the cusp of yet another technological revolution with the advent of AI. Between heightened US-China tensions, the west at a standoff with Russia and conflict in the Middle East deepening fractures in the multilateral system, the climate for global agreements is less than favourable.

A recent first meeting between US and Chinese officials in Geneva on the security risks of AI could have led some to believe in a thawing of tensions at least on the tech issue. But Kurbalija tempers expectations. While the two rivals have shown a willingness to talk to one other over AI, other fields of technology remain “the scene of geopolitical battles”, as seen with the latest tit-for-tat bans relating to US microprocessors.

“More countries are losing hope in multilateral cooperation and preferring digital sovereignty to protect their digital space from the spillover from global geopolitics,” he said.

Developing countries fear that AI risks widening the gap with wealthy nations. The concerns were palpable at a conference on AI and development organised by the Central African Republic and the International Organisation of la Francophonie (OIF) last week, where Tunisian ambassador Sabri Bachtobji compared it to the Covid pandemic – countries were all asked to share data about the spread of the virus in their territories, but when vaccines were developed, they were left on their own.

The comments come as developing countries demand for equity at fraught negotiations for a pandemic treaty set to conclude at the World Health Assembly this week.

Can New York provide guidance?

As the layers of challenges add up, the Global Digital Compact set to be adopted in New York in September is being touted by some as the much-needed guiding star. However, observers are wary of vague proposals for new bodies and new funds included in the draft.

“The biggest concern is how to avoid duplications because there are bodies which are already discussing digital issues,” said Kurbalija, cautioning against undermining Geneva’s already well-established ecosystem of technical institutions, including the International Union of Telecommunications and the World Meteorological Organization but also its orbiting satellites such as WSIS, the GIF and the Human Rights Council.

“We often discuss in abstract what should be digital or AI governance. In the lines of the famous architecture Bauhaus principle, form should follow the function. What do we really want to regulate with AI? Long-term risks? Short-term risks? Education? There is no clarity yet, but I hope that the compact can provide that in the discussion around governance,” said Kurbalija.

Mehdi Snene, senior advisor of the UN’s tech envoy, argued at the AI and development meeting that governance discussions should focus on the outputs rather than the technology itself, for example anticipating intellectual property rights implications of AI-produced content.

One of the proposals in the text is to create an IPCC-style scientific panel for AI and emerging technologies that could help governments understand the potential and risks of the technology and perhaps even provide policy guidance – a “reasonable compromise”, according to Kurbalija. He is sceptical that countries would agree on creating a new regulating body.

“Small and developing countries are tired of this proliferation of bodies and the politicisation. They’re looking for practical solutions,” he said.

Read more: AI transparency: where should the goalpost be planted?

Funding is another issue where disagreement is likely to continue. The current draft proposes the establishment of a $100 million Global Fund for AI and emerging technologies for Sustainable Development in 2025 built on voluntary donations from public, private and philanthropic sources. Speaking at the conference on AI and development, OIF representative Henri Monceu pleaded in favour of such a fund as a critical piece to make sure developing countries also reap the benefits of AI.

“There is a need to support the development of digitalisation and AI in developing countries, but I'm not sure that it will fly,” said Kurbalija. He points to a proposal at the last WSIS meeting in Tunis in 2005 to create a digital solidarity fund, ultimately rejected by developed countries such as the United States and the United Kingdom. Developed countries are often the ones expected to foot the bill for such funds.

But even if the prospects for major political decisions look slim, Kurbalija notes it shouldn’t be a reason to bring everything to a halt. “While everybody's busy discussing what should be the framework for the future of AI governance, Geneva organisations should start contributing to digital AI governance through their activities”, which he noted some organisations are already embracing.




Please login to post a comment.