Competition among major powers over artificial intelligence (AI) has gained significance in recent years. The 18th BRICS Summit, held in New Delhi on September 12–13, 2026, demonstrated that this competition also has implications for the BRICS countries. During the summit, Chinese President Xi Jinping proposed that China take the lead in establishing a "BRICS AI Open-Source Community". The proposal was part of a broader set of initiatives that included a digital ecosystem cloud platform, cooperation in scientific and technological innovation, and closer links between AI technologies and smart manufacturing.

The New Delhi Declaration, dated September 12, adopted a more cautious approach to AI. It called for greater access to AI resources while emphasising safety, security, inclusiveness and reliability. It also committed BRICS members to implementing the BRICS Leaders' Statement on the Global Governance of Artificial Intelligence. However, the declaration did not explicitly endorse Xi's proposed open-source community. The gap between China's proposal and the declaration's broader language illustrates the limits of consensus within BRICS on emerging technologies.

China's Vision for AI Cooperation Within BRICS

China sees BRICS as a platform for shaping an international AI ecosystem in which multiple global powers can participate. During the summit, Xi presented an "open source and inclusive AI initiative", under which Beijing would promote cooperation in developing and applying large language models (LLMs), organise specialised AI seminars and training courses, and help build an open AI ecosystem.

The initiative reflects China's growing role in the development of open-weight AI models. Chinese companies have released prominent models, including Alibaba's Qwen series and DeepSeek's R1. Such models can offer developers greater flexibility to adapt and deploy AI systems, although the degree of openness, licensing conditions and access to training data vary between models.

The New Delhi Declaration, however, did not explicitly endorse China's proposed community. Instead, it emphasised shared priorities, including wider access to AI resources, safety, security, inclusiveness and reliability. It also highlighted the importance of international cooperation in AI science and innovation, the development of trustworthy AI technologies, and the mitigation of potential risks.

This language reflects the consensus-based nature of BRICS. Rather than committing the group to a specific institutional arrangement proposed by one member, the declaration established broad principles for cooperation. Its approach leaves individual members considerable room to pursue their own AI priorities.

China's Open-Source AI Agenda and the Global South

China's open-source AI initiative is directed in part at the Global South. Beijing has presented open-source and inclusive AI as a means of strengthening technological capabilities in developing economies. Greater access to AI models can allow countries to adapt technologies to local languages and needs, reduce dependence on proprietary services, and develop domestic expertise.

However, access to model weights alone does not guarantee technological sovereignty. Countries also need computing infrastructure, skilled personnel, suitable data, financing and effective governance frameworks to develop and deploy AI systems independently.

At the World AI Conference and High-Level Meeting on Global AI Governance in Shanghai in July 2026, China continued to promote international cooperation on AI. Its broader approach has included training opportunities and engagement with regional and multilateral organisations, including ASEAN, the African Union, the Arab League and the Shanghai Cooperation Organisation.

For developing countries, the appeal of open-weight models lies in the possibility of adapting AI to local requirements without relying entirely on expensive proprietary platforms. Nevertheless, the practical benefits will depend on the availability of computing resources, the terms under which models can be used, and the capacity to manage data security and other risks.

India's Position and the Diversity of BRICS Interests

India's position helps explain the cautious language of the New Delhi Declaration. The IndiaAI Mission has focused on strengthening domestic AI capabilities, including computing infrastructure, indigenous foundation models and skilled talent. At the India AI Impact Summit in February 2026, India also emphasised making AI accessible and using it to address real-world challenges across sectors.

New Delhi has sought to advance equitable access to technology and greater representation for developing countries in global AI governance. Its approach involves engagement with multiple partners rather than exclusive dependence on either Washington or Beijing. China's open-source proposal may therefore offer opportunities for cooperation, but participation in a Chinese-led initiative would not necessarily align with every aspect of India's technological and strategic priorities.

Other BRICS members also have different interests. The United Arab Emirates, for example, has developed its own Falcon family of AI models through the Technology Innovation Institute, while also pursuing international technology partnerships. Its experience illustrates how a country can invest in domestic AI capabilities while engaging with external technology providers.

Russia's priorities include technological self-reliance and the development of domestic AI capabilities, particularly in areas considered strategically important. Brazil and South Africa, meanwhile, have reasons to emphasise AI's potential contribution to development, public services, environmental applications and skills creation. These priorities are not mutually exclusive, but they illustrate the different economic and strategic considerations that shape members' approaches to AI.

Consequently, the absence of an explicit endorsement of China's proposed community should not automatically be interpreted as evidence of a concrete disagreement. It may instead reflect the diversity of national interests and the difficulty of translating broad support for AI cooperation into a common institutional framework.

China's BRICS Chairmanship in 2027

China will assume the BRICS chairmanship in 2027 and host the 19th BRICS Summit, giving Beijing an opportunity to pursue its AI cooperation proposals further. As chair, China could seek to build consensus around practical collaboration, including joint research, LLM development, technical training and pilot projects.

The extent to which these proposals advance will depend on the willingness of other members to participate, the resources they are prepared to commit, and the arrangements adopted for intellectual property, data protection, technical standards and governance. Cooperation on specific projects may prove more feasible than agreement on a single, overarching AI framework.

Overall, the difference between Xi's New Delhi proposal and the language of the declaration illustrates the distinction between China's ambitions for AI cooperation and the level of agreement achieved among BRICS members. The declaration did not reject China's AI diplomacy; it simply did not explicitly endorse the proposed open-source community.

BRICS members may benefit from access to Chinese open-weight models while continuing to develop their own AI capabilities and maintain partnerships with other technology providers. However, the extent of such cooperation will depend on national priorities, infrastructure, security considerations and the conditions attached to individual models and partnerships.

The New Delhi outcome therefore points to the possibility of practical cooperation without a fully integrated BRICS AI ecosystem. Whether China's proposal develops into a broader multilateral initiative will depend on its ability to address the varied interests of BRICS members and translate general principles into arrangements that they are willing to support.

(The author is a student in the Department of Political Science at North Bengal St Xavier's College, Rajganj, India. Her academic interests include the growing role of artificial intelligence in international relations and global politics. The views expressed are personal. She can be contacted at liza.paul2007@gmail.com)