Aligning India’s AI future with climate goals

This article is authored by Angelina Chamuah and Lianne D’Souza.

At the recently concluded Conference of Parties (COP)29, a part of the discourse centred on the growing role of digital technologies, particularly Artificial Intelligence (AI), in addressing the climate crisis. While AI offers potential for climate action, from optimising renewable energy grids to enhancing climate modelling, its environmental footprint is also substantial and growing. The recent COP29 Declaration on Green Digital Action, announced on the first ever “digitalisation day” at COP, reflects a growing international recognition of these trade-offs – highlighting the need to align AI trajectories with climate goals. For India, this conversation takes on a particular urgency, as the nation advances its AI ambitions, while striving to meet its climate targets under the Paris Agreement.

Artificial Intelligence

India has set ambitious climate targets, including reducing emissions intensity by 45% from 2005 levels, achieving 50% non-fossil power capacity by 2030, and attaining net-zero emissions by 2070. It has made notable progress, with renewable energy now accounting for 46.3% of India’s total capacity, and an 85% cut in fossil fuel subsidies through its ‘remove, target and shift’ strategy. However, significant challenges persist, especially in transitioning from coal, which remains critical to meeting the country's growing energy needs.

Alongside its climate ambitions, India is also rapidly advancing its AI ecosystem. Since launching the National Strategy for AI in 2018, and more recently the IndiaAI Mission, the country has been focussing on strengthening its AI capabilities--including enhanced compute infrastructure, developing Large Multimodal Models (LMMs), and a strong, indigenous AI ecosystem for innovation. As part of the mission, the Government has also earmarked ₹5,000 crore to provide over 10,000 Graphics Processing Units (GPUs). Additionally, three Centres of Excellence (CoEs) are being established to foster AI innovation in critical sectors such as health care, agriculture, and sustainable cities. Ranked 14th globally in data centres, India is also poised for significant expansion in data infrastructure. With India’s AI ambitions on the rise, critical challenges around environmental sustainability need closer attention.

Data centres, which serve as the backbone of AI infrastructure, are both energy-intensive and water-reliant, contributing to emissions and local resource strain. At present, data centres account for nearly 1% of the global Greenhouse Gas (GHG) emissions. For a country grappling with water scarcity and the need to decarbonise its energy systems, these pressures present tangible challenges. Moreover, the hardware supply chains underpinning AI—ranging from semiconductors to GPUs—are linked to extractive industries that generate significant ecological and social costs. This is compounded by the exponentially growing computational demands of AI, which double roughly every 100 days, intensifying emissions and resource pressures. Furthermore, research suggests that e-waste generated through generative AI processes could reach a total of 1.2-5.0 million metric tonnes by 2030.

Given these dynamics, it is clear that AI policy and climate policy cannot operate in silos. Without deliberate strategies to address these challenges, the environmental costs of AI could offset its potential benefits for climate action. AI policy is increasingly climate policy, and both must co-evolve to ensure that technological advancements support, rather than hinder, environmental sustainability.

Five specific measures can help advance this agenda.

• First, India's progress in renewable energy, especially solar power, offers a major opportunity to green the AI infrastructure. While much research focuses on developing energy-efficient AI models, there is also a need to look at the bigger picture. This means considering the entire system that supports AI, such as adopting policies incentivising renewable-powered data centres.
• Second, embedding circular economy principles in the AI lifecycle is crucial. Policies promoting the recycling, reuse, and responsible disposal of electronic components can mitigate the environmental impact of AI hardware.
• Third, prioritising AI applications in climate adaptation and resilience—such as early warning systems for extreme weather events, and optimised water resource management — can not only directly support India’s climate targets, but also offset the environmental implications of AI infrastructure. While initiatives like Mission Mausam are a step in the right direction, involving local communities and integrating indigenous knowledge in these efforts is also a must.
• Fourth, establishing energy-efficiency benchmarks for AI training, deployment, and operations is vital. For example, Article 40 of the European Union (EU) Act, 2024 outlines a framework for creating harmonised standards designed to enhance the resource performance of high-risk AI systems, with a strong emphasis on energy efficiency. India could adopt similar principles to set industry standards.
• Fifth, while principles of safety and reliability, inclusivity and non-discrimination, equality, and privacy feature firmly in NITI Aayog’s Responsible AI strategy, sustainability must also be looked at as a core principle of responsible AI practices.

Ultimately, addressing the intersections of AI and climate challenges requires moving beyond piecemeal interventions, and taking a holistic account to how we align AI development and governance with sustainability principles. Sustainability is not an attribute of individual technologies or policies, but emerges from the broader socio-technical systems in which these technologies operate. As a nation committed to both technological and environmental progress, aligning these dual imperatives is essential for achieving a sustainable digital future.

This article is authored by Angelina Chamuah, programme director, Future Foresight, Transitions Research and Lianne D’Souza, associate researcher and assistant professor of law, NLSIU, Bengaluru.