Andreas Hoepner, co‑leader of the GreenWatch team at University College Dublin and head of the EU Platform on Sustainable Finance’s Data Science Hub, argues that AI’s net impact on climate may be positive. While many warn that large‑language models spike electricity use, Hoepner says the technology’s ability to accelerate battery research outweighs the energy cost. He notes renewable‑sourced data centres curb impact.

Current neural‑network LLMs excel at rapid binary screening, a strength Hoepner cites for combing through tens of millions of potential battery materials. In the past three years, hyperscale firms partnered with labs to speed material discovery a hundredfold, and DeepMind’s protein‑folding breakthrough that earned a 2024 chemistry Nobel illustrates AI’s capacity to solve complex scientific puzzles. Such speed could halve R&D cycles industry‑wide.

The trade‑off remains energy‑intensive: LLMs consume power comparable to average data‑center mixes, and many generate incorrect emissions data, wasting resources. Yet Hoepner remains optimistic that a breakthrough akin to DeepMind’s could compress battery development timelines enough to meet the Paris Agreement’s 2050 electrification goal. Investors should watch AI‑driven material platforms as the next frontier for clean‑tech financing. Funding pipelines already reflect this emerging demand.