AI, a huge opportunity and serious risk for climate: UNEP official

Artificial Intelligence offers significant potential to help address the climate crisis, but its high energy and water demands could worsen environmental challenges if not managed properly, a senior official of the United Nations Environment Programme (UNEP) has said. In an interview with PTI, Martin Krause, Director of the Climate Change Division at UNEP, described AI from a climate‑policy perspective as both a powerful opportunity and a serious risk.
According to the PTI report, Krause highlighted that AI‑driven early weather‑warning systems can predict floods and storms with much greater accuracy and, if deployed at scale, may protect hundreds of millions of people. He also pointed to AI’s role in integrating renewable energy into national grids—such as in India—where the technology helps recalibrate and balance power flows, easing the integration of intermittent solar and wind generation. Satellite‑based AI tools are also being used to detect and track emission sources more promptly, supporting faster responses to climate‑relevant events.
At the same time, Krause flagged concerns about the large water and energy footprint of AI‑driven data centres, which support the compute‑intensive models powering generative AI and advanced analytics. He emphasised that AI‑infrastructure expansion must be coupled with a shift to renewable energy to avoid increasing greenhouse‑gas emissions from fossil‑fuel‑driven power sources. Site‑specific assessments, he said, are needed to determine the right mix of energy sources for data‑centre clusters, ensuring that AI benefits do not come at the cost of depleting scarce natural resources.
The UNEP official noted that the United Nations environment policy apparatus has begun to treat AI as a distinct environmental issue, reflected in a dedicated resolution on the sustainable use of AI adopted by the United Nations Environment Assembly in 2025. In line with earlier global discussions, UN Secretary‑General António Guterres has also urged that AI data centres switch to clean‑energy sources instead of shifting environmental and social costs onto vulnerable communities.