Artificial intelligence (AI) is a double-edged sword in the fight against climate change. On one hand, it holds great potential to cut down carbon emissions. On the other, its current path points to increased energy use, according to the International Energy Agency (IEA). While AI could eventually outpace the energy needs of expanding data centers, we must address their environmental impact now.
Sam Altman, CEO of OpenAI, is optimistic about AI’s potential to tackle climate change. Yet, today’s data centers, mostly powered by natural gas, are major contributors to greenhouse gas emissions. This has sparked calls for new gas plants and even the revival of old coal facilities to meet rising energy demands.
Cleaner energy options like geothermal, nuclear, and hydroelectric power are available, but they’re costly and take time to deploy. This idea that AI will offset future emissions is similar to carbon offsets—where current pollution is balanced by future projects. It’s a bit like putting off dealing with a problem until later.
The IEA points out that AI is already being used in practical ways, such as detecting methane leaks and improving energy efficiency. AI also shows promise in advancing materials science and enhancing climate science and disaster response. Despite these benefits, the IEA warns that the impact of AI on emissions by 2035 is still uncertain.
Under the best conditions, AI could cut emissions by 1.4 billion tons by 2035—three times the emissions from AI data centers in the most hopeful scenario. But this depends on major tech advances and supportive policies. Without the right incentives, AI might end up helping fossil fuel exploration more than reducing emissions.
The IEA emphasizes the unpredictable role of AI in cutting emissions. There’s currently no momentum ensuring these technologies will be widely adopted. Climate change is urgent, with global temperatures nearing critical levels, and the window for effective policy is closing.
As industries aim for net-zero emissions by mid-century, the continued reliance on fossil fuel infrastructure is a big hurdle. Even if AI eventually offsets its emissions, the damage done in the meantime will linger in the atmosphere for centuries. It’s crucial for the AI sector to adopt sustainable energy solutions now, rather than waiting for future tech breakthroughs.
Some tech companies are already leading the way by prioritizing renewable energy sources, building solar farms, and reviving nuclear plants. But these efforts need to become the norm to prevent worsening the climate crisis. The time to act is now, as delaying only increases the environmental costs.
