Five Challenges with Multistakeholder Initiatives on AI

Sep 15, 2021

Applications of artificial intelligence are creating new spaces that need governance. Multistakeholder initiatives are promising, but they must meaningfully include everyone whose voices need to be heard. Experience so far in digital governance shows that meaningful inclusion is a serious challenge.

Multistakeholder initiatives have often worked well when they have narrow objectives in setting technical standards, the Internet Task Force being an example. It was natural to borrow the model to try to handle broader issues around ethics, rights, and development – but this brings a new challenge of different stakeholders bringing different political and ideological paradigms.

Experience so far suggests that efforts at meaningful inclusivity in multistakeholder governance fall short in five key ways.

1. Not all stakeholders are created equal.

Multistakeholder initiatives typically include four types of stakeholders – international organizations, governments, private sector, and civil society. However, inviting stakeholders from each category can become a box-ticking exercise rather than a genuine attempt to bring in diverse perspectives from different groupings around the world.

Supposedly global initiatives often comprise only Western nations, leaving out lower-income countries – and typically also China, despite it being a tech giant. From the private sector, only big tech players tend to be invited or spotlighted. Smaller tech companies meanwhile, and those which deal with digital technologies only tangentially, rarely have their voices heard, even though they may have more nuanced insights based on their niche focus areas or markets served. Civil society representatives are usually large, international NGOs with offices in capital cities, rather than the kind of local actors who may better understand what is happening on the ground.

2. Participation is resource-intensive.

Even if many of the typically excluded groups mentioned above – local civil society organizations, smaller companies, and lower-income countries – are invited to participate, in practice they may not have sufficient time, money or capacity.

Part of the problem is the sheer number of initiatives in tech governance. There are around 600 alone in soft law mechanisms, many of which will invoke multistakeholderism. Participating in multiple such initiatives quickly stretches an organization’s resources; indeed, it can be a full-time job even to work out which initiatives it should prioritize taking part in.

3. Money shapes the agenda.

How and by whom initiatives are convened remains a big challenge in multistakeholder tech governance. What legitimacy does the convenor of an initiative have? How do they define the issues to be addressed, and decide which actors they approach to participate?

Also: from where do those actors get their funding? If a private sector player is funding governments and civil society to participate in a multistakeholder initiative, they can hardly be expected to deviate robustly from that player’s interests. So while in theory an initiative can look like it brings in a heterogeneity of perspectives, in practice there can be multiple actors all pushing the same line.

4. They can be talking shops.

It is often not clearly articulated whether multistakeholder initiatives are intended to be merely consultative or capable of enforcing decisions. The Internet Governance Forum, for example, is widely seen as a mere talking shop.

When multistakeholder arrangements lack the power to enforce, it enables the bigger players in government and the private sector to engage in forum-shopping or retreat into more exclusive arrangements, like Davos, to take the real decisions.

5. The relationship of multistakeholderism to multilateralism is unclear.

Many nations that are not typically included in multistakeholder arrangements would prefer a multilateral approach to tech governance. In part this is because they are familiar with the multilateral means of engagement, easing the resource challenges of navigating multistakeholder initiatives.

They may also feel also that multilateral approaches bestow more legitimacy. All stakeholders, in their way, can claim to be representative of people – governments because they are elected; civil society organizations because they are funded by donations; and private sector players because they depend on giving consumers what they want – but, ultimately, it is states that have the authority to implement and enforce governance outcomes.

In reality, both multistakeholderism and multilateralism have a role to play in tech governance. The challenge is figuring out how they should fit together.

Nanjira Sambuli is a researcher, policy analyst, and advocacy strategist interested in and working on understanding the unfolding impacts of ICT adoption and how those impact governance, media, entrepreneurship, and culture, with a keen focus on gender implications.

You may also like

A Dangerous Master book cover. CREDIT: Sentient Publications.

APR 18, 2024 Article

A Dangerous Master: Welcome to the World of Emerging Technologies

In this preface to the paperback edition of his book "A Dangerous Master," Wendell Wallach discusses breakthroughs and ethical issues in AI and emerging technologies.

MAR 22, 2024 Podcast

Two Core Issues in the Governance of AI, with Elizabeth Seger

In this "Artificial Intelligence & Equality" podcast, Carnegie-Uehiro Fellow Wendell Wallach and Demos' Elizabeth Seger discuss how to make generative AI safe and democratic.

FEB 23, 2024 Article

What Do We Mean When We Talk About "AI Democratization"?

With numerous parties calling for "AI democratization," Elizabeth Seger, director of the CASM digital policy research hub at Demos, discusses four meanings of the term.