BRICS 2026 Delhi Declaration puts AI at heart of Global South growth

For the AI and machine learning community, the BRICS Delhi Declaration signals a major shift in the geopolitical landscape of model development and deployment. The bloc's emphasis on energy efficiency and inclusive access could shape architecture choices, infrastructure investments, and data-governance norms across a market of billions, while its governance commitments create new interoperability and compliance considerations for AI builders.

On September 12, 2026, BRICS member nations released the New Delhi Declaration, a sprawling communiqué that elevates artificial intelligence from a peripheral technology issue to a central organising principle for the bloc's cooperation across economic growth, healthcare, energy and development. According to the ANI-distributed text of the declaration, the grouping recognised that AI 'presents immense opportunity to stimulate economic growth and sustainable development for all,' while simultaneously calling for greater international cooperation to improve access to AI resources, promote trustworthy AI technologies, foster AI science and innovation, and mitigate potential risks. This dual framing—AI as both engine of growth and source of governance challenges—places BRICS firmly within the intensifying global contest to shape norms for frontier technologies.

“

For the AI and machine learning community, the BRICS Delhi Declaration signals a major shift in the geopolitical landscape of model development and deployment.

The declaration's emphasis on the Global South is particularly significant. BRICS, which has expanded beyond its original five members in recent years, is positioning itself as a counterweight to Western-led AI governance efforts. By stressing 'a just, fair, equitable and prosperous future of all countries, especially those of the Global South,' the bloc is signalling that it intends to make AI access and capacity-building a central plank of South-South cooperation. The language on promoting energy efficiency of AI systems also reflects growing concern within developing economies that the computational demands of large AI models could deepen existing energy inequalities. This is not merely rhetorical: energy-intensive AI infrastructure has become a flashpoint in global climate and industrial policy, and BRICS members such as India and China are actively building domestic AI compute capacity while seeking to manage electricity constraints.

The declaration explicitly ties AI to BRICS' economic, industrial and financial cooperation. It highlights the work of the BRICS Centre for Industrial Competencies in supporting manufacturing companies, including small and medium-sized enterprises, in adopting Industry 4.0 and advanced digital production technologies. This signals a shift from high-level statements toward the operational layer of AI diffusion—factory automation, supply chain intelligence, predictive maintenance—across member economies. For AI developers and enterprise technology providers, the BRICS framework could become a structuring force for data governance, interoperability standards and procurement priorities in a bloc representing over 40 percent of the world's population.

The declaration also institutionalises the governance agenda by committing to implement the BRICS Leaders' Statement on the Global Governance of Artificial Intelligence. It congratulates India on the February 2026 AI Impact Summit and acknowledges the World AI Conference and High-Level Meeting on Global AI Governance held in Shanghai. These references suggest an emerging division of labour within BRICS: India as a convenor of value-oriented AI summits, China as host of technical and governance conferences, and the bloc as a whole as a venue for harmonising positions ahead of multilateral forums such as the United Nations. The absence of specific technical standards or enforcement mechanisms in the declaration text, however, means the follow-through will depend on concrete working groups, funding commitments and regulatory alignment that have yet to be demonstrated.