Microsoft is reportedly reassessing its 2030 commitment to match 100% of its hourly electricity consumption with renewable energy purchases, as the scale and energy intensity of AI infrastructure, particularly for Azure and Copilot, outpace earlier sustainability assumptions.
The review, reported by Bloomberg and Reuters, comes amid industry-wide expanded data center buildouts tied to AI demand.
The possible shift underscores a structural tension between AI infrastructure expansion and clean energy procurement at scale. Hyperscalers are deploying multi-gigawatt data centers to support large model training and inference workloads, materially increasing baseline electricity demand. A single gigawatt, now a relevant unit in hyperscale planning, can power approximately 750,000 to a million homes for an hour, depending on region and conditions, illustrating the magnitude of consumption associated with AI services.
This has driven diversification in energy sourcing strategies across the sector, including increased reliance on nuclear and natural gas to ensure a consistent baseload supply. Microsoft itself entered into an agreement with Constellation Energy in 2024 to restart a unit at the Three Mile Island Nuclear Plant.
Microsoft’s earlier climate targets were set prior to the current acceleration in generative AI adoption. No final decision on Microsoft’s 2030 target has been made.