COP30: AI as a Climate Justice Ally or the New Face of the Extractivist Model?

During RIO-92, for the first time in a large United Nations summit, there was a network of computers connected to the internet, allowing the dissemination of environmental debates in real time. The provision of an internet connection for the event was due to the participation of civil society, especially in Latin America, in its advocacy for digital technology in the service of human rights. The event also played. a jey role in boosting the adoption of the internet in Brazil.

Three decades later, this link between digital technology and environmental concerns, especially involving artificial intelligence (AI), is even more manifest. First, for the growing scientific evidence of the alarming impact of AI on the environment. Second, by the centrality of the subject in global geopolitics, demonstrated by the race for technological development, reflected in conflicts over minerals, such as illegal mining in the Congo, disputes for access to Brazilian indigenous lands, and even mining rights for Ukrainian rare-earth minerals figuring into the negotiations for the end of the war between Russia and Ukraine.

With 2024 surpassing the 1.5°C global warming threshould, discussions – not new, but highly relevant in the cotext of AI - have highlighted the dual role of these technologies in climate change. For example, AI can be an ally in environmental conservation monitoring, but also increases energy consumption and the extraction of natural resources. Studies such as one by the International Energy Agency (IEA) show the growing demand in the sector, predicting that, after consuming an estimated global amount of 460 terawatt hours (TWh) in 2022, the total power consumption of data centers may exceed 1,000 TWh in 2026, the approximate electrical consumption of Japan. Similarly, projections from the University of California, Riverside, indicate that the consumption of water related to AI processing may reach 4.2 to 6.6 billion cubic meters by 2027, equivalent to the annual water usage of four to six Denmarks.

For the development and use of AI to be truly aligned with the principles of socio-environmental justice, it is fundamental that laws and public policies take on such questions as energy efficiency, technological autonomy, mitigation of environmental impacts, inclusive governance and robust accountability mechanisms. It is necessary to recognize that the current model of production and technological innovation frequently ignores ecological and social limits and are based upon digital extractivism, concentration of power and technological dependence. The very idea of “sustainable development” has been, many times, devoid of meaning, serving only to justify practices that continue to degrade the environment.

In its presidency of COP30, Brazil will be able to play a fundamental role in directing discussion toward clarifying the environmental consequences of AI and strengthening the development of technologies to mitigate its impacts. Brazil has unique comparative advantages, such as biodiversity, cultural diversity and social innovation, which should be central to this agenda. The relationship of indigenous people and traditional communities with nature, sustainable agricultural practices, generation of renewable energy and the fight for green cities, are all strategic elements that may highlight Brazilian participation, especially at COP30.

The Brazilian Approach

Formally, Brazil featured the intersection between AI and the environment in its Nationally Determined Contribution (NDC) to the United Nations Framework Convention on Climate Change (UNFCCC), citing the Brazilian Plan for Artificial Intelligence (PBIA). Furthermore, the Ministry of Foreign Affairs, in partnership with the Alexandre Gusmão Foundation, promoted the seminar “Artificial Intelligence and the Climate” in January.

With this initiative, Brazil arrives at the COP30 in a strategic position, building on its presidency of the G20 in 2024 and of the BRICS in 2025, strengthening the opportunity to influence global debates in an uncertain geopolitical scenario. In its domestic plan, the Brazilian government has signaled its approach toward AI and the environment, which, however, prioritizes immediate economic opportunity, without regard to socio-environmental impacts. Some initiatives stand out:

• Brazilian Plan for Artificial Intelligence (2024-2028): Structured on five axes, sets a budget of R$23 billion through 2028, with R$500 million designated to the Program for Sustainability and Renewable Energy for AI and R$2.3 billion for the so-called “green data centers.” It is expected that such resources will inspire the creation of “sustainable” data centers and AI infrastructure, prioritizing renewable resources and highly efficient electrical energy technologies. Despite seeking to reinforce technological autonomy and making mention of sustainability, the Plan lacks robust strategies to address the environmental impacts of AI;
• Regulation of AI (PL 2338/2023): Approved in the Senate in December of 2024, this bill provides, through investments, the expansion of data centers and promotion of research and infrastructure. Although the bill discusses energy efficiency and certification programs to mitigate environmental impacts, its approach focuses solely on the promotion of technological development;
• National Policy for Data Centers (PNDC) – MP 1.318/2025 (REdata regime): In September 2025, the federal government enacted a Provisional Measure establishing the National Policy for Data Centers and the Special Tax Regime for Data Center Services (REdata). The measure provides significant fiscal incentives, including suspension of federal taxes on equipment acquisition, while requiring the use of renewable or “clean” energy sources, high water efficiency standards (WUE ≤ 0.05 L/kWh), allocation of at least 10% of installed capacity to the domestic market (or equivalent investment in R&D), and a 2% reinvestment obligation in national research projects. Although these provisions appear to incorporate sustainability concerns, key terms such as “clean energy” and other safeguards remain undefined and dependent on future regulation. Civil society organizations, including LAPIN, IDEC, and IP.rec, have criticized the policy for its lack of transparency, absence of prior public consultation, vague socio-environmental commitments, and omission of issues such as e-waste management and prior consultation with indigenous and traditional communities.

These examples demonstrate the necessity of amplifying the discussion about AI and sustainability beyond development agendas. The lack of a more critical debate over the environmental and social risks limits the potential of these initiatives. The main example that we present here is the case of data centers, which, although often labeled “green” or “sustainable,” demand special attention to their true environmental and social consequences.

Paths toward initiatives that integrate socio-environmental challenges

So that Brazil may develop strategies that align with its territorial necessities while contributing to global efforts to mitigate climate change, it is essential that its policies take a more holistic approach, such as:

1. Ensure energy transition and efficient use of natural resources: Digital infrastructure cannot, for example, depend on fossil fuels and must follow rigorous criteria of energy and natural-resource consumption;
2. Map out the socio-environmental impacts of AI in different contexts and beyond energy infrastructure: It is necessary to identify the impacts upon various sectors, such as agriculture, where AI could affect land use, biodiversity and the way of life of local communities, requiring specific safeguards;
3. Strengthen informed participation and dialogue in decision making: COP30 and other forums should not be limited to multilateralism. It is fundamental that access to information be guaranteed, and that traditional peoples and communities, academia and members of civil society be included in the processes of decision making to ensure that solutions are aligned with local realities and the protection of rights;
4. Require transparency of environmental impacts: It is essential that companies and institutions be required to divulge the environmental impacts of AI infrastructure, including data centers and associated support chains. Evaluations of algorithmic impact should be mandatory and address sustainability and socio-environmental impacts;
5. Strengthen global partnerships: Brazil must seek strategic alliances with other nations affected by climate change, to influence international decisions and guarantee funds for sustainable, socially responsible solutions, including debates on the COP Technology Implementation Program (TIP).

Conclusion

In the context of the global climate crisis, the presence of the COP30 in Brazil confers not only international visibility, but a concrete responsibility to the country in the face of disputed digital and energy transitions. It is essential that the socio-environmental impacts of technological advances be confronted critically, especially in the face of unchecked expansion of large-footprint digital infrastructure like data centers—frequently associated with the intensive use of energy and natural resources, and installed in territories marked by inequality.

There is no longer any space, under the pretext of “national sovereignty” or “economic development,” for public policies related to artificial intelligence and the environment to advance without consideration for criteria such as transparency, social participation, inclusive governance and accountability. The COP30 should take the spotlight as a strategic space to build an agenda in which digitalization and energy transition go hand in hand, guided by human rights, climate justice and collective responsibility.