TEDIC, via the MAP.IA coalition, sets out a Global South civil society perspective on AI governance

TEDIC, writing as a member of the MAP.IA coalition, sets out Global South civil society positions ahead of the first UN Global Dialogue on AI Governance, scheduled for July. The coalition is led by the Centre for Communication Governance at National Law University Delhi together with the Global Network Initiative. TEDIC says it also belongs to the Global South Alliance and the Global Coalition for Tech Justice, and that it submitted detailed recommendations to the UN consultation.

The piece starts from a tension between governance frameworks that are weak and non-binding and the pace of deployment, which deepens inequality between North and South. The authors say the AI summit series from Bletchley to Paris has been shaped by corporate and Northern government interests, that civil society roles remain largely symbolic, and that declarations from Paris in 2025 and India in 2026 did not bind states. They describe Latin America as caught in data extraction and technological dependence.

Four tensions are identified: the legitimacy of multistakeholder processes, uneven agenda-setting power, the gap between declared principles and implementation, and the question of what AI is built for. The authors argue that risk-based frameworks should fit local conditions rather than copy the EU AI Act, and they criticize regulatory sandboxes. They call for safety across the whole lifecycle, moratoria on rights-incompatible systems, and interoperability anchored in human rights. They add that language models are weak in Spanish and Portuguese and nearly absent for over 400 Indigenous languages, and they want free, prior and informed consent. Further demands include funding for local evaluation, protection from retaliation, human rights impact assessments, mandatory independent audits of high-risk systems and strategic litigation.