AI and Climate Goals Converge at COP30: Navigating the Energy Paradox of Digital Transformation

How AI can accelerate the energy transition, rather than compete with it

Starting with COP30, collaboration is essential to ensure that AI development accelerates the broader energy transition. The COP30 climate summit in Belém, Brazil is happening at a pivotal moment for the global clean energy transition. But the event has also attracted noticeably limited participation from major players including the US, Russia, India and China, underscoring growing geopolitical and implementation challenges. Now more than ever, the countries gathered for COP30 face pressure to move from ambition to action, advancing the commitments made at COP28 to triple global renewable energy capacity and double energy efficiency by 2030. COP30 has also placed digital technologies and artificial intelligence (AI) firmly on the agenda, recognising AI's energy paradox. While this technology could accelerate clean energy deployment and system optimization, its rapidly rising need for electricity poses new challenges for grids, policy frameworks and long-term planning. This paradox is likely to persist as AI-driven demand surges. Data centre investments are projected to reach around $1.1 trillion by 2029, alongside a global push to scale renewables and upgrade grid infrastructure. AI's growth has created a powerful incentive for technology firms to invest in clean, reliable power. But it also raises a critical question for COP30 negotiators: What if the new clean energy capacity being built remains siloed, powering AI but not strengthening the wider energy systems that communities, industries and households rely on? Stronger coordination between technology leaders, utilities and policy-makers is essential to ensure AI's growth supports, rather than competes with, global decarbonisation goals. Data centres could represent around 3% of global electricity demand in 2030, raising concerns about the ability to support this demand while also meeting global net-zero commitments. AI's rapid expansion has forced technology firms to confront the limits of existing energy systems. Data centres and the advanced chips that drive AI are among the fastest-growing sources of global electricity demand. By 2035, data centres in the US alone could account for 8.6% of total electricity use - more than double their current share. Globally, data centres consumed around 415 TWh in 2024 and this is expected to more than double by 2030, according to the International Energy Agency (IEA). In response, companies such as Microsoft, Amazon and Google are signing long-term clean-power contracts, investing directly in generation projects and even financing early-stage nuclear and geothermal ventures. These investments could have positive effects, helping to de-risk innovative clean technologies, demonstrate their commercial viability and build local clean energy capacity. However, they also expose a structural tension as AI's clean energy leadership is driven primarily by self-supply imperatives, rather than by system-level planning that delivers shared benefits. Consequently, much of the new capacity being developed may not substantially support the broader electrification of transport, heavy industry or communities. And as tech giants secure long-term renewable deals, smaller players risk being priced out. This could create a two-speed transition that would undermine equitable decarbonisation. Several US utilities, for example, are currently delaying fossil-plant retirements and building new gas facilities to meet surging data-centre demand. If this kind of growth continues, it could strain grid reliability and slow the wider transition to clean power. AI-driven energy investment should be channelled in ways that strengthen entire systems, not just individual data centres. As electricity systems face growing pressures from vehicle and building electrification, reindustrialisation and population-driven demand, the rapid rise of AI introduces an additional layer of complexity. The IEA estimates that data-centre growth could account for more than 20% of total power demand growth in advanced economies through 2030. If most new generation capacity is channelled into powering AI workloads, fewer resources may remain for hard-to-decarbonise sectors, slowing the broader energy transition. At COP30, discussions of "twin transitions" (the convergence of AI and the energy transition) highlight how these forces could work together to drive growth, energy security and climate action. The launch of the AI Climate Institute during COP30 reflects this ambition, positioning AI as a tool for empowerment, particularly for the Global South, where access to clean, affordable energy remains critical. The clean energy capacity powering AI must also reinforce grids and expand energy access to benefit society as a whole. This balance is essential to align technological progress with climate ambition. AI's potential impact on the energy transition could extend well beyond simply consuming electricity. It could serve as a key enabler of system intelligence, improving renewable forecasting, grid balancing and predictive maintenance. AI could also optimise building efficiency and enable flexible demand that adjusts to variable solar and wind output. When applied intentionally, AI could transform the energy system, making it more adaptive, resilient and equitable, rather than simply adding to demand. Achieving this balance will require system-level partnerships across sectors. Based on the latest cross-industry insights from the Forum's AI Energy Impact initiative, four areas for collaboration stand out: Such steps would help ensure that AI's clean energy momentum strengthens public systems, rather than creating parallel energy ecosystems available only to the largest firms. To drive collaborative progress on energy intelligence, future COPs should incorporate several key themes. Dedicated sessions on "AI & clean energy infrastructure" would ensure issues like renewable contract lock-in, equitable access, grid integration and scaling replicable models are formally on the agenda. By amplifying regional case studies such as green data centres, integrated renewable capacity for AI and shared grid access, other countries can learn and replicate these developments. Finally, embedded policy and governance dialogues, covering standards for "green compute" and clean energy for AI data-hubs, would help participants ensure that infrastructure benefits communities, not just hyperscalers. AI's growing electricity demand is both a challenge and an opportunity. It can accelerate the commercialization of clean technologies and reshape energy systems for the better - but only if the benefits extend beyond data centre needs. AI is already driving the next phase of energy innovation. The challenge now is to make sure this innovation strengthens the entire energy system, powering not only the algorithms of the future but also the societies they aim to serve.

A.I., Data Centers and Climate Impact: How Leaders Are Rewriting the Energy Equation

The environmental footprint of A.I. and data centers is under increasing scrutiny. Concerns over rising electricity demand, water use and the carbon cost of compute now dominate headlines and policy discussions. Yet a deeper, data-driven look reveals a far more nuanced -- and more promising -- story: A.I. and data centers can indeed coexist with climate goals and will, in fact, lead the global march toward meeting them. Sign Up For Our Daily Newsletter Sign Up Thank you for signing up! By clicking submit, you agree to our <a href="http://observermedia.com/terms">terms of service</a> and acknowledge we may use your information to send you emails, product samples, and promotions on this website and other properties. You can opt out anytime. See all of our newsletters A resilient data infrastructure is foundational to modern economic development. Nations that invest in robust data centers and A.I. ecosystems gain the capacity to transition to digital-first economies, which are inherently more efficient and less carbon-intensive. Data centers do not operate in isolation to serve their own needs. Rather, they power the intelligence layer that helps every major industry -- from logistics and manufacturing to healthcare and finance -- operate with greater precision, automation and resource efficiency. In this way, the benefits multiply far beyond the facilities themselves. One of the clearest ways to evaluate greenhouse gas (GHG) performance is to examine how much economic output a nation generates for each ton of emissions released. The International Data Center Authority (IDCA) uses a measure of metric tons of GHG per million U.S. dollars of economic output (or nominal GDP). The global average today sits at 357 tons per million dollars. The United States, home to the world's densest concentration of data centers, operates at roughly half that level. Several E.U. nations, with particularly strong, modern digital infrastructure, perform even better, and the most efficient Nordic economies produce emissions at nearly twice the efficiency of the U.S. By contrast, the least-efficient economies tend to be those dominated by heavy industry or agriculture. China and India, for example, generate more than twice the global average. Many underdeveloped nations across Africa and Asia also produce high emissions relative to their economic output due to limited technological modernization and slower transitions away from carbon-intensive sectors. This is not to say that the United States is an exemplar of best practices in addressing GHGs. It remains below the global average in renewable energy adoption, is heavily reliant on gasoline- and diesel-powered transportation and continues to struggle with a government that oscillates between ambivalence and hostility towards addressing emissions reduction. Moreover, the U.S. and wealthier E.U. nations have outsourced much of their high-emissions manufacturing -- their "dirty work" -- to China, India and less-developed nations, complicating global accounting and underscoring that no country can claim moral high ground. Still, the data highlights an essential point: if the entire world operated at the U.S. level of economic efficiency, global emissions would fall by roughly half. And if the U.S. itself moved closer to the efficiency levels of leading E.U. nations, the reductions would be more significant. Achieving that scale of improvement requires confronting three central questions: How can the world's largest emissions producers work together to address reductions across heavy industry? China, the U.S., India and Russia, along with industrial powerhouses like Japan, South Korea and major petrostates, must work in coordination to decarbonize heavy industry. A.I. can be a powerful lever here, optimizing automated manufacturing, refining supply chains and enabling predictive efficiency improvements at scale. Equally important is accelerating the shift toward low-carbon construction materials, particularly alternatives to cement and steel, which account for roughly 15 percent of global GHG emissions. How can developing nations build their digital economies with minimal climate impact? Developing nations are the subject of many discussions at COP30, the annual United Nations meeting focused on climate change abatement, this year held in Belém, Brazil. Most use only 2 to 5 percent of the electricity consumed per person in the developed economies, and their data center infrastructure remains even less mature than their electricity grids. IDCA's research shows a strong correlation between digital infrastructure and socioeconomic advancement, making sustainable grid development and low-carbon construction paramount. These countries have a narrow but critical opportunity to build modern digital economies without inheriting the emissions burdens of previous industrial revolutions. How can the next generation of massive A.I. centers being planned and built today prioritize emissions reduction? As thousands of new A.I. facilities are planned and constructed worldwide, their environmental profiles will continue to transform the global economy into a digital economy. These centers must be designed for maximum efficiency, powered by increasingly clean grids and built with clear sustainability criteria. But their impact won't end at their footprint. Strong A.I. backbones can accelerate global decarbonization by enabling smarter transportation networks, optimizing energy systems, transforming healthcare analytics, advancing meteorological modeling and empowering scientific discovery across environmental domains. If world leaders, business executives and citizens collectively demand that A.I. and data centers be aimed at humanity's most pressing scientific and environmental challenges, then A.I. centers will be understood not as climate liabilities, but as essential climate tools. The path forward will not be simple, but it is one that we are fully capable of navigating. Mehdi Paryavi is the founder and CEO of the International Data Center Authority (IDCA), the world's leading Digital Economy think tank.