A vision for democratizing AI emerged at the heart of Paris this week, as voices from the Global Majority and other regions challenged the traditional power structures shaping our technological future. Digital Action and Forum on Information & Democracy co-hosted an official side event at the AI Action Summit, on “Building Equitable AI: Global Voices for Information Integrity,” where experts converged on the need to reimagine how artificial intelligence can serve humanity's diverse needs.
The AI Commons Initiative, presented by our Campaigns and Partnership Director for Latin America, Vladimir Cortés, emerged as a framework for democratizing AI through global citizen power. Its four-pillar approach — establishing AI Equity Lab Networks, Citizens' Design Councils, a Policy Innovation Lab, and Multi-stakeholder Oversight mechanisms — presents a practical roadmap for ensuring technology serves all communities equitably.
The rationale for adopting such new solutions were explored through a look at major urgent problems experienced across different countries. Carla Vreche from Conectas Brazil revealed the harsh realities of AI misuse in politics, particularly through deepfake technologies targeting women candidates. She cited the unauthorized use of Brazilian children's images for AI training and the discriminatory impact of facial recognition systems leading to racial profiling, including a troubling case of mistaken identity at a football match in Sergipe. Beyond these concerns, Vreche detailed Brazil's complex journey toward AI regulation, where 46 bills have been introduced in the past years. A significant breakthrough came with Senate Bill 2338's approval in December 2024, though the victory was bittersweet. While the bill establishes risk-based governance rules, intense corporate lobbying led to weakened provisions for social participation and worker protections, with facial recognition systems remaining largely unrestricted.
Michael Oghia of the Global Forum for Media Development highlights how media organizations spend between 20-30% of their resources on cloud computing in an increasingly concentrated market, revealing an infrastructure crisis facing the Global Majority. This stark economic reality underscored the pressing need for more accessible and diverse AI ecosystems.
Camille Grenier of the Forum on Information and Democracy presented a groundbreaking solution: a voluntary certification mechanism for public interest AI. This innovative approach aims to bridge the gap between users and deployers while creating economic incentives for responsible AI development, particularly crucial in regions lacking comprehensive regulatory frameworks.
Rodrigo Durán from Chile's National Center for Artificial Intelligence brought forward insights on balancing innovation with accountability, particularly regarding data access and intellectual property rights. He highlighted the need to embrace collaboration and scale initiatives to the regional level, emphasizing the importance of cross-border partnerships in shaping AI development.
The dialogue emphasized pragmatic solutions and swift action in response to rapid technological change. Participants stressed the importance of breaking down silos between environmental and technological initiatives, fostering regional collaboration, and ensuring that diverse values guide AI development. We are delighted that our gathering could mark a step toward ensuring that the Global Majority's voices shape the future of AI governance.
Comments