As artificial intelligence (AI) accelerates its grip on global industries and sectors, the question of control becomes impossible to ignore. Who decides how AI evolves? Whose interests are being prioritized? While corporations and governments drive much of the conversation, civil society remains a crucial but often sidelined voice in AI decision-making. This imbalance - reflects historical legacies and risk perpetuating “digital colonialism”, were at the heart of discussions at a recent webinar on AI and the Global South, where experts and civil society organizations argued that AI governance must be more inclusive—or risk deepening existing inequalities. The webinar was a deep dive into a recent research study by CARE International and Accenture, amplifying voices from 12 countries across Africa, Asia, the Middle East, and South America. The report, AI and the Global South: Exploring the Role of Civil Society in AI Decision-Making, highlights key perspectives on AI’s risks, opportunities, and governance gaps. 

A Race Without Rules? 

"AI is moving fast, but we need to ask: who is shaping it, and who is benefiting the most?" Mika Välitalo of Fingo posed this question at the start of the discussion, framing a debate that underscored how current AI development is largely dictated by the so-called Global North. The event, featuring contributions from Fingo, CARE International, Accenture, the regional coalition from Central African Repongac, the Forus network and various civil society organizations (CSOs), painted a risky picture: the voices of local communities remain largely absent in AI governance, and the consequences of this exclusion are profound. 

John Carruthers of Accenture, acknowledged AI’s transformative potential but warned that without localized insights, much of this potential will be wasted. "If AI development is not guided by local experiences, its social impact will be superficial at best and harmful at worst," he cautioned. His concerns were echoed by multiple speakers, who argued that AI must reflect diverse cultural and economic realities to be truly transformative. 

The Civil Society Challenge: four key pathways CSOs can push for more inclusive AI governance. 

The discussion highlighted four key pathways CSOs can push for more inclusive AI governance, which were discussed in group discussions and collaborative tools. As one CSO participant in the study stated, “Every community and every country has its own clear difficulties, and because of that it makes us have certain kinds of customised intersectional programs.” Still, evaluating pathways by considering diverse stakeholders’ perspectives helps uncover commonalities and generate forward-looking considerations to inform action. 

1. AI Literacy & Cross-Sectoral Knowledge Sharing – "Many communities are interacting with AI without fully understanding its risks and opportunities," noted Suzy Madigan of CARE International. The absence of multilingual resources and accessible education on AI risks widens the knowledge gap, limiting communities' ability to advocate for their needs. Improving AI literacy among communities and CSOs is essential for informed decision-making, whether individuals are using AI tools or affected by AI-driven decisions. At the same time, technologists must better understand the contexts and lived experiences of Global South communities to design AI models that consider their real-world impacts, especially in humanitarian and development settings. “There is a manipulation of the vulnerability... they (communities) are signed up to things that they don't know about or are not fully informed about,” explained a CSO participant to the study in South Africa. 

2. Increasing local decision-making and representation across the AI lifecycle – AI governance tends to be centralized, ignoring local expertise. "The idea of localization is well understood in development work, but it's missing in AI governance," observed Fergus Gleeson of Accenture’s Human Sciences Studio. This disconnect leaves communities without a say in the technologies that shape their daily lives. AI applications are growing rapidly, but civil society’s role in decision-making remains limited. CSOs, with their deep community ties, can help identify local challenges and foster engagement. However, effective participation requires addressing issues like trust, problem identification, and practical engagement strategies. 

3. Strengthening advocacy on the contextualised impacts and desired outcomes of AI – Many CSOs lack the funding and research capacity to effectively engage in AI policy discussions. Projects such as the Civil Society Alliances for Digital Empowerment (CADE) - EU-funded consortium of nine organizations – are now working to strengthen civil society organizations in internet governance. The project promotes inclusive participation, particularly from the Global South, and advocates for an open, free, and secure internet by supporting net neutrality, privacy, and freedom of expression.  

4. Improved digital infrastructure and equitable data governance – AI models often rely on data that fails to represent Global South realities, leading to biased outcomes. Ahmed Al-Khatib from the Yemen-based Mwatana for Human Rights warned, "If AI is trained on unrepresentative data, it will misinform and perpetuate biases." The need for better rural connectivity and digital infrastructure was a recurring theme. Improving AI’s societal impact requires strong digital infrastructure and quality data. Addressing data and digital divides in the Global South is crucial, ensuring communities have agency over their data, not just as sources or users. 

Can Civil Society Work With Big Tech? 

A point of discussion was the relationship between CSOs and multinational technology firms. A representative from a multinational tech company in the study, acknowledged the ethical concerns: "We want to work with CSOs to develop ideas together, but we struggle with the perception that we are being extractive." This tension between cooperation and resistance remains a central challenge for CSOs seeking influence in AI governance. 

Suzy Madigan offered a pragmatic take: "Many inside these companies want AI to be ethical, but they need external pressure to push for change internally." The implication is clear—CSOs must find allies within the industry while maintaining a critical stance. 

Opportunities and Challenges of AI in Africa – Repongac 

Repongac, Forus regional coalition in Central Africa presented perspectives on the need for civil society’s involvement in decision-making to ensure that AI development aligns with local needs. While several African countries have introduced initiatives to strengthen local capacities and promote innovation, their implementation varies due to differences in resources and infrastructure. 

AI presents significant opportunities in Africa, including improvements in public services, local innovation, community empowerment, and global economic participation. In sectors such as healthcare, education, agriculture, and public administration, AI-driven solutions can provide locally adapted interventions. For example, the E-Farm startup has developed sensors to measure soil aridity, addressing a challenge unique to African agriculture. AI is also supporting entrepreneurship, particularly among young people, and enabling local communities to manage resources, prevent disasters, and enhance civic participation. 

Several obstacles hurt AI adoption in Africa, such as the lack of regulatory frameworks, limited local participation in AI governance, and a shortage of skills in AI-related fields. Additionally, inequalities in digital access and the absence of locally relevant data further complicate the widespread implementation of AI, preventing many African communities from fully benefiting from its potential. 

Strategies for AI Governance in Central Africa 

To strengthen AI governance and digital inclusion, Repongac focuses on building local capacities through education and training, international cooperation and equitable partnerships, and creating coalitions to amplify civil society's voice. It also promotes transparency and accountability in AI policies, advocates for regulatory frameworks that protect citizens' rights, and works to raise public awareness about the benefits and risks of AI. 

Public education campaigns are central to Repongac’s strategy. It produces educational content, organizes public dialogues, and pressures international organizations and tech companies to include more African representatives in AI decision-making forums. 

AI and the Future of Advocacy 

One thing is certain: civil society can no longer afford to be a passive observer in AI governance. The stakes are too high. AI can be a tool for empowerment, but without proactive intervention, it will only reinforce existing inequalities. 

"AI literacy isn’t just about understanding technology; it’s about ensuring communities have agency in their digital lives," Gleeson concluded. The future of AI is being written now, and civil society must demand a seat at the table before it’s too late. 

The full report, AI and the Global South: Exploring the Role of Civil Society in AI Decision Making, is available online and serves as a roadmap for policymakers, technologists, and civil society leaders committed to a fairer AI future. 

Read more on Forus work on Digital Governance