Geneva AI Week: Did meaningful engagement improve mere tokenism?
By Raymond Amumpaire
August, 2026
From 07–10 July 2026, Geneva, Switzerland, saw a convergence of conversations for civil society and governments on global AI governance, from the inaugural UN Global Dialogue on AI (UNGD), a mechanism established by the Global Digital Compact, to the World Summit on the Information Society (WSIS) and the International Telecommunications Union (ITU) AI for Good Global Summit. The week-long engagement marked progress in global AI governance as AI technologies continue to shape the lives of current and future generations.
Expectations of stakeholders ahead of the summit
Global AI governance is steadily acquiring momentum through institutions, processes and spaces through which the future of this frontier technology can be negotiated. Despite the foregoing, a survey conducted by the author about AI week across participants from civil society, academia, the private sector and the technical community representing Africa, the Arab Middle East and Asia-Pacific revealed four connected concerns: a) the gap between participation and influence; b) the absence of concrete pathways from dialogue to implementation; c) the growing attention to the link between sovereignty and local economic value; and d) the need for governance processes to continue between major international meetings.
83% of the respondents entered the week expecting concrete governance commitments and greater Global Majority influence. As compared to a space to learn about emerging tech or even observing technical demonstrations, they hoped to see governance, international coordination, influence, and institutional outcomes. Other expectations bordered on practical AI solutions, human rights and child safety protections, international standards, and funding and partnerships. Post-AI week, 66% of respondents rated the week as partially meeting these expectations, 18% as slightly meeting them, and 16% as not meeting them at all.
Why did expectations fall short? The responses allude less to dissatisfaction with dialogue itself than to uncertainty about what the dialogue produced. Survey findings point to participation being tokenistic and influence from the Global Majority remaining symbolic. Also, while the structured nature of conversations at the UN was welcome, there remained no intersessional work, leaving AI to be governed by a narrow group of technologically powerful companies and states. Earlier this year, the Group of 77 and China warned that a number of issues could impede developing countries from fully participating in and benefiting from AI development.
Important questions underlined conversations, but still need further exploration: Do Global AI governance conversations really respond to Global Majority issues? What played out in Geneva left many wondering what influence Africa and the Global Majority have in the greater scheme of AI governance.
One would wonder how much meaningful influence Africa and the Global Majority had over the agenda and outcomes at the AI week? And, does being invited into global governance spaces translate into having power within them? It is important to note that some participants from Africa/Global Majority evidently encountered opportunities to influence discussions. But in all fairness, even the three minutes allotted during cluster sessions for submissions were insufficient.
The UNGD was initially premised on expanding participation through a common platform that included all UN member states and non-state stakeholders. Initiatives such as the Global Majority AI Governance Fellowship by the Heinrich Böll Stiftung supported a number of persons from the constituency to participate in the processes. Others include communities such as the Multistakeholder Approaches to Participation in AI Governance (MAP-AI) and Participatory AI Research & Practice Symposium (PAIRS), which convened cross-learning sessions aimed at Global Majority participation.
However, in practical terms, the absence of meaningful engagement and participation is obviously most critical for African countries in the Global Majority because while the continent is missing an opportunity to shape global AI governance, it continues to make it a neocolonialistic source of minerals that power the development of these technologies, labour that can be exploited every which way and reduces the youthful continent to just another market that lines the pockets of the Western tech bro-ligarchy.
Questions about sovereignty
Important for me were the discussions around digital sovereignty and how economic value is morphing in an increasingly AI-dependent world. This is central to what AI governance primarily means for African countries and communities. Development and economic value, human rights, safety and accountability, protection against surveillance or manipulation, and better public services took centre stage, suggesting that AI governance ought to be holistically viewed as compared to being framed as a question of only regulating risk.
As attested to by the outcome of the side-event discussion at the Diplo head office in Geneva under the auspices of the Civil Society Alliances for Digital Empowerment (CADE) project (co-funded by the EU; opinions expressed are those of the authors and do not represent positions of the EU), these issues are central to understanding where power lies and values flow.
Sustainability
Discussants agreed on the need to pay greater attention to the highly under-addressed issue of infrastructure that powers data centres and its implications for energy, water, land and emissions. Discussions also encouraged reflection on the labour costs of these models, shedding light on labour violations in Kenya, where content moderators sued Sama Source. This third party sourced labour that powered the Meta AI machine.
On the brighter side of things, the International Labour Organisation (ILO) touted its recently adopted Decent Work in the Platform Economy Convention, 2026 (No. 193), which marked a milestone in the development of international labour standards dealing with digitisation. The convention promotes decent work opportunities and measures to facilitate formalisation, and establishes protections for digital platform workers.
Human rights
While rights and safety were important topics of the week, and there was agreement as to the centrality of respect for human rights in AI governance, there was a disappointing lack of discussion and consensus in the room about broader human rights issues. For example, there is an increasing proliferation of AI in warfare that is being used and abused by both state and non-state actors. A case in point is the use of Palantir Technologies, Amazon Web Services (AWS) and Google, Intel, OpenAI and Microsoft technologies in the genocide in the State of Palestine at the behest of Israel. This is substantiated in the Report of the Special Rapporteur on the situation of human rights in the Palestinian territories occupied since 1967, A/HRC/59/23: From economy of occupation to economy of genocide.
Looking forward
As the world looks ahead to the next segment of the UNGD happening in New York (May 2027), critical questions remain: a) Who has meaningful influence within these new institutions? b) What does AI sovereignty mean for countries that do not control much of the infrastructure on which AI depends? c) How should governance account for the environmental and social costs of the infrastructure boom? d) Can states address AI accountability without addressing concentrated corporate power? and perhaps most importantly, e) How does a global governance process move from successful convening to sustained implementation? How do these questions resonate with issues now appearing more explicitly in international policy discourse?
To begin to answer these critical questions would require collectively evaluating and reaching consensus on whether these spaces, albeit risking duplication, have evolved to create the much-needed impact in this area. In turn, such exercises will necessarily require:
- The practical implementation of existing global-level platforms and initiatives to ensure that the Global Majority are meaningfully engaged in priority area selection, agenda- setting, partaking in deliberations, as well as collective brainstorming exercises geared towards human rights-oriented, inclusive, and tailored AI governance models which are cognisant of their contexts
- The embedding of follow-up on commitments made on global platforms and feedback mechanisms, in which the views and concerns of the Global Majority are part and parcel of such initiatives, thus shaping policies and decisions even after the closure dates of summits and conferences. In parallel, good practices can be embedded within and across initiatives and platforms at the global level to ensure their sustainability from one generation to the next.
- The imperative to open up existing spaces to the Global Majority, to enable their voices (especially young and underrepresented leaders) to be heard in practical terms, thus to enable them to influence the nature and trajectory of AI-related dialogue, alongside key policy and decision-making processes at the global level. At its backdrop, such access also necessitates the alignment of existing platforms (to avoid parallel conversations and siloed efforts) and actions geared towards ensuring the ownership of both conversations and processes by all those who engage, including those from the Global Majority.
Such embedding of meaningful participation and engagement of the Global Majority will ripple further onto regional and sub-regional initiatives. For Africa, this could drive the amplification of meaningful outcomes at the continental level. This will require further, more serious implementation of efforts to move from tokenism to meaningful participation, and a willingness to address challenging issues.
Civil Society Alliances for Digital Empowerment (CADE) is a project co-funded by the European Union. The views expressed in this blog are those of the author and do not necessarily reflect the views of the European Union or CADE.
