At Keenious, we are committed to minimising our environmental footprint. We recognise that artificial intelligence has a real environmental impact, from energy-intensive data processing to water use and carbon emissions. With global AI energy demand projected to increase tenfold, we are dedicated to transparency and actively working to reduce energy consumption and greenhouse gas emissions [1].
📊 Energy & Water Use
Each document processed through Keenious consumes up to about 0.33 watt-hours, roughly 0.35% of the energy it takes to heat a cup of tea. That's around a 50% reduction on our 2025 estimate, driven by infrastructure efficiency improvements. Our infrastructure, hosted across Google Cloud and AWS, also draws water for cooling, which we actively monitor and mitigate through our deployment choices, including a Belgian data centre that draws non-potable canal water rather than drinking water.
🌱 Carbon Emissions
Our cloud infrastructure across Google Cloud and AWS generated approximately 4.7 tonnes of CO₂e (market-based) for June 2025 to May 2026, a figure that reflects our choice of sustainable cloud partners. But cloud is only part of the picture. Once we include Scope 3 emissions, indirect impacts like business travel and commuting, our total footprint for the year is around 54 tonnes of CO₂e, equivalent to the annual emissions of roughly three average Americans.
We report Scope 3 separately as it forms the larger share of our overall impact.
If you want to learn more about these figures, check out our Estimated Energy Consumption of a Keenious Search 2026.
🔄 Sustainable by Design
We are aiming to design Keenious in the most environmentally responsible way possible. To achieve this goal, we:
Locate workloads on lower-carbon grids wherever possible. We've migrated our Google Cloud workloads to France and Belgium, away from Ireland's higher-carbon grid.
Optimise our models to reduce redundant processing and select fit-for-purpose models rather than maximum-power ones.
90% of our team works remotely. Remote work has been shown to cut commuting emissions by up to 54%.[3]
Continuously monitor and report our environmental metrics transparently.
🌍 Want to help shape a greener future for AI? Let us know how we can improve at contact@keenious.com.
[1] International Energy Agency. (2024). Electricity 2024: Analysis and forecast
to 2026. IEA.
2118a/Electricity2024-Analysisandforecastto2026.pdf
[2] Note calculations are based on market-based methods (MBM). For more information on MBM see here. All equivalencies are calculated using the United States, Environmental Protection Agencies Greenhouse Gas Equivalencies Calculator - https://www.epa.gov/energy/greenhouse-gas-equivalencies-calculator#results
[3]Tao, Y., Yang, L., Jaffe, S., Amini, F., Bergen, P., Hecht, B., & You, F. (2023).
Climate mitigation potentials of teleworking are sensitive to changes in
lifestyle and workplace rather than ICT usage. Proceedings of the National
Academy of Sciences, 120(39), e2304099120.