AI energy use
Artificial grey cells run on grey electricity.
And nobody knows how much. We make it visible — with an open methodology and every figure clearly labelled.
A blind spot that is growing fast
Vehicle fleets, buildings and business travel are accounted for in detail. The energy behind AI use? Not in any balance sheet, any standard, any number — and it is growing faster than anything else.
Our approach
Measure
Capture usage through the providers' official interfaces — or via export and manual entry where no interface exists.
Disclose
Every energy factor with its source and date. What is measured is called measured. What is modelled is called estimated.
Name the uncertainty
We show ranges instead of point values. Where our model ends, we say so.
3
companies already measure their AI use here — ourselves included.
Measure
We measure ourselves first
Degeler Consulting has been tracking its own AI footprint for months: energy, CO₂ and water of our own usage, updated daily and open to anyone. Using the same factors we publish here — and the same ranges wherever we do not know more precisely.
See our own footprintQuestions about the methodology, or interested in a pilot? Write to us.
hello@degeler.consulting