BRICS Information Centre
|
BRICS Information Centre |
|
BRICS Leaders' Statement on the Global Governance of Artificial Intelligence
Rio de Janeiro, Brazil, July 6, 2025
[pdf]
Artificial Intelligence represents a milestone opportunity to boost development towards a more equitable future, fostering innovation, enhancing productivity, advancing sustainable practices, and concretely improving the lives of people everywhere on the planet.
To achieve that goal, global governance of AI should mitigate potential risks and address the needs of all countries, especially those of the Global South. It must operate under national regulatory frameworks and the UN Charter, respect sovereignty as well as be representative, development-oriented, accessible, inclusive, dynamic, responsive, grounded in personal data protection, the rights and interests of humanity, safety, transparency, sustainability, and conducive to overcoming the growing digital and data divides, within and between countries. A collective global effort is needed to establish an AI governance that upholds our shared values, addresses risks, builds trust, and ensures broad and inclusive international collaboration and access, including capacity building for developing countries, with the United Nations at its core.
The proliferation of governance initiatives and the diverging views in multilateral coordination at the international level may aggravate existing asymmetries and the legitimacy gap of global governance on digital matters, further eroding multilateralism as a result.
To support a constructive debate towards a more balanced approach, we, the BRICS Leaders, have agreed on a set of guidelines to foster responsible development, deployment, and use of AI technologies for sustainable development and inclusive growth. These guidelines, that strictly refer to the use of AI in the non-military domain, should be applied through either domestic or applicable international frameworks, as well as through the development of interoperable standards and protocols, in inclusive, transparent, and consensus-based processes.
The United Nations is Central to Global AI Governance. Fragmentation and duplication of global AI governance efforts must be avoided at all costs. We must strengthen AI international governance through the United Nations system as a fully inclusive and representative international framework. We must promote emerging markets and developing countries’ (EMDCs) and the Global South’s meaningful participation and role in decision-making processes under UN initiatives, while recognizing the complementary roles of regional and multi-stakeholder frameworks, and encourage policy exchanges and dialogues on AI, with a view to spur innovation and promote economic growth.
Collaborative Governance of AI is Complex, but Possible. A network of relevant stakeholders, from developed and developing countries, should provide contributions with unique expertise, perspectives, and resources, within their respective roles and responsibilities. Recalling the fundamental and leading role of governments in conducting AI governance, we will closely cooperate with the private sector, civil society organizations, international organizations, the technical and academic communities, and all other relevant stakeholders towards inclusivity and representation.
Digital Sovereignty and the Right to Development are Central to Global AI Governance. We firmly support the right of all countries to harness the benefits of the digital economy and emerging technologies, particularly Artificial Intelligence, while upholding fundamental rights, to establish their own regulatory frameworks within their jurisdictions, to develop capacities in AI research, foster technological autonomy and innovation, ensure data protection, and promote their own digital economy, by enhancing digital infrastructure, fostering local talent, and ensuring safety and security of their citizens from AI risks.
Fair Competition and Market Regulation are at the Heart of an Equitable AI Future. The digital economy requires safeguarding the rights and obligations of States, companies, and users under domestic legal frameworks and regulations and applicable international agreements to level the playing field towards innovation and economic growth. We underline the importance of avoiding regulatory fragmentation and of promoting fair and transparent market regulations that encourage competition, enhance competitiveness and avoid distortions of any kind, thus ensuring a sustainable and healthy market environment.
Data Governance is Key to Inclusive AI Governance. We highlight that fair, inclusive and equitable governance of data is critical to enable developing countries to harness the benefits of the digital economy and emerging technologies, including artificial intelligence. We also acknowledge the need to establish data governance frameworks, in accordance with our commitments under the BRICS Data Economy Governance Understanding, to build trust and to confer equitable, agreed upon, safe, and secure access to quality data by developing countries, while respecting applicable legal frameworks related to the right to privacy, protection of personal data, algorithmic transparency, intellectual property rights, and national security.
Access to AI Technology Should be Fair, Equitable, Enabling and Inclusive. All countries, irrespective of their stage of economic development, have the right to benefit from, develop, and use AI. We underline international cooperation to facilitate access to AI technology and critical components, to remove barriers to financial resources required for AI research and innovation, and to develop necessary knowledge, skills, and risk management frameworks to effectively leverage AI technologies, especially in low and middle-income countries.
A Balanced Approach is Needed to Protect Intellectual Property and Safeguard the Public Interest. We need a balance between proprietary rights, transparency, and accountability to safeguard public interest, the international transfer of technology, and compliance with domestic legislation and applicable international law. Appropriate protection of intellectual property rights and in particular copyright against unauthorized AI use must be in place to prevent exploitative data extraction and violation of privacy, allowing for fair remuneration mechanisms. Safeguards must ensure accountability and compliance with relevant legislation, including transparency over AI model inputs and outputs.
AI Should be Open and Promote Innovation Ecosystems. We encourage open-source development and international scientific and technological cooperation, in line with national policies and priorities, through Open Science and Open Innovation mechanisms as key enablers to building capacities in AI research, development, innovation, data protection, data sovereignty, and deployment, as well as to allow researchers, developers, and organizations to scrutinize, audit, and contribute to safe, secure, trusted, and transparent AI systems. We should ensure meaningful participation and foster inclusive collaborative approaches to avoid barriers to global AI development and supply chains. We encourage the development of resource-efficient, specialized, and open foundational models to promote the growth of innovation ecosystems in AI.
International Standards Should Lead to Inclusive, Representative and Accessible AI. The development of technical standards, specifications, and protocols for AI systems must involve the public sector, standards development organizations, and specialized UN agencies to ensure trust, interoperability, safety, and reliability of AI technologies across the development process and different platforms and applications. We must avoid standard setting processes being used as barriers to market entry for smaller companies and developing economies.
AI must be Available to All. Robust infrastructure, meaningful connectivity, and digital inclusion are prerequisites for the deployment of AI in any country. Digital government services, including those based on digital public infrastructure, may act as catalysts to inclusion in the digital economy and access to citizen’s rights and benefits. Prevalent asymmetries in capacities and infrastructure should be addressed by domestic measures, including multilateral development banks (MDBs), to advance local adoption and development of AI in EMDCs.
AI can Lead to Collaborative Sustainable Development. We seek to support AI applications, including through open source, that address critical development challenges in sectors such as healthcare, education, security, transport, energy, agriculture, environment, water resources, and waste management, in accordance with national laws and priorities. The benefits of AI for sustainable development should be continuously promoted. We will focus on research, development, and innovation initiatives that foster and empower local technological capacities and aspirations and bridge technological gaps between developed and developing countries.
Environmental Sustainability is a Prerequisite. AI should be used to mitigate emissions, adapt to climate change, enhance environmental conservation efforts, promote sustainable resource management and optimization, maintain ecological balance, and contribute to the evolution of the social and economic pillars of sustainable development, in line with national priorities. AI development and deployment should be sustainable and should minimize their own environmental impact, addressing aspects such as greenhouse gas emissions, energy and water consumption, use of materials, and electronic waste. The windows of opportunities opened by the transition to a low carbon emissions economy should allow developing countries to harness the momentum and to lead through technological advancements in AI.
AI Must Ensure Decent Work and Enhanced Productivity. AI systems have the potential to enhance productivity, stimulate innovation, and create job opportunities, but they also pose challenges, concerns, and risks over working conditions, work intensity, job displacement, and threats to employment and workers’ dignity. It is imperative to safeguard the rights and wellbeing of all workers, particularly those directly affected by the digital transformation, considering the rapid growth of digital platforms and the increasing structural impact of AI, including generative AI, on the job market. It is also paramount to ensure that the design, deployment, and use of AI take into consideration adaptability and compatibility in terms of human resource capabilities and utilize the advantages of technology, such as AI, to promote decent work and full and productive employment.
AI Should be a Tool to Transform Human Education and Learning. As AI systems curate information, personalize experiences, and offer advice, we are concerned that excessive reliance on these systems may affect cognition, decision- making processes, and the ability to navigate complex information landscapes. We affirm our disposition to foster digital literacy skills, in particular of teachers and learners, to critically evaluate AI-generated content, understand embedded biases within algorithms, and cultivate a strong sense of intellectual autonomy and critical thinking skills.
AI Must be Inclusive. We recognize risks relating to misappropriation and misrepresentation of knowledge, heritage, and cultural values that are insufficiently represented in datasets and AI models. We reaffirm the importance of ethical, transparent, and accountable frameworks, such as UNESCO’s Recommendation on the Ethics of Artificial Intelligence. AI systems must be developed with rigorous respect to the linguistic, cultural, racial, geographical, and demographic diversity of all countries. This includes international cooperation for training on comprehensive, multilingual, and inclusive quality datasets and the development of local AI talent.
Discriminatory Bias must be Mitigated. We need robust tools to identify and mitigate errors and negative algorithmic biases and ensure independent audit mechanisms that safeguard fairness, provide standards on assessment of risks from bias, and avoid discrimination and exclusion. Women, minorities, persons with disabilities, and groups in vulnerable situations, such as children, youth, and the elderly, tend to be the most affected by AI systems trained with biased materials. Interdisciplinary collaboration among people from different backgrounds is key to establishing standards, enhancing explainability of model operation and outputs, and creating practical strategies that guard against negative bias and support the responsible and equitable development of AI systems.
Public Interest Comes First. Under a human-centered approach, we will foster a harmonious human-machine relationship, with ultimate authority and oversight from humans, where AI continues to be built as a powerful tool for augmenting human capabilities. We will prioritize human supervision, transparency in AI decision- making, and effective accountability mechanisms so that AI technologies can be developed and deployed responsibly and securely, mitigating risks and maximizing societal benefits.
Stick to the Facts. The generation of realistic fake text, images, audio, and video content poses significant threats to information integrity and authenticity and may lead to manipulation of public opinion, incitement of social unrest, and the undermining of trust in public institutions. We will seek a multi-faceted approach to promote information integrity, greater centrality to media education strategies, and local communication efforts. This includes developing tools to quickly flag disinformation and misinformation, fostering digital literacy and critical skills in individuals to better evaluate online content, and establishing clear ethical guidelines and regulations, in line with the need for protection of privacy and digital data and for the development and use of AI technologies in information dissemination.
Safety and Trust must be Inherent to AI Systems. We acknowledge the need to achieve safe, ethical, trustworthy, and responsible AI development for the benefit of all. Countries must collectively promote innovation in and access to AI technologies, while addressing both immediate and long-term risks in accordance with national policies and security considerations. We reaffirm the importance of collectively addressing and protecting against the risks posed by malicious uses of AI technologies. AI systems, as well as other information and communications technologies, should be carefully designed in a way that enables effective detection and prevention of misuse, such as fraud, computer attacks, cybercrimes, or data manipulation.
A Prudent Approach is Needed Towards AGI. It is crucial that research on Artificial General Intelligence (AGI) be developed ethically and deployed responsibly and in a trusted manner, with a view to helping accelerate economic growth, especially in EMDCs, and address pressing economic and social challenges. If concentrated in the hands of a few actors, AGI could exacerbate inequalities and create new forms of technological dependence, raising serious challenges to sustainable development.
Promoting a Fair and Inclusive AI Landscape. BRICS countries will seek to adopt a common, proactive position to advance an equitable digital environment for all, identify common perspectives and intensify global collective efforts towards operationalizing the UN Global Dialogue on AI Governance and the International Independent Scientific Panel on AI, while ensuring meaningful roles and participation of developing countries. We will coordinate efforts to promote our shared approaches to AI, including these guidelines, across relevant international fora. We welcome contributions to further develop these guidelines, particularly from other developing countries, and will remain open to revisiting them.
Source: Official website of Brazil's BRICS presidency