Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Microsoft’s reported water use rose from 6.4 million cubic meters in 2022 to 7.8 million in 2023, while its reported greenhouse-gas emissions grew from about 12 million metric tons in 2020 to about 15 million in 2023. The increases coincided with the company’s expansion of cloud and AI infrastructure, but they do not show how much was caused by AI workloads alone. The figures, reported in a May 17, 2024 Futurism article, describe a broader corporate footprint.
What rose, and by how much?
The figures reported by Futurism imply that Microsoft’s water use increased by about 22% between 2022 and 2023. Its rounded emissions figures imply an increase of about a quarter from 2020 to 2023. The original coverage’s reference to roughly 15 metric tons is missing the word “million”; the intended scale is approximately 15 million metric tons.
These are reported corporate totals, not a measurement of the footprint of one AI product, datacenter, or workload. The cited coverage does not provide enough detail to break the totals into water withdrawal versus consumption, individual water sources, or emissions scopes. It is therefore not possible from these figures alone to say precisely how much came from facility operations, purchased electricity, construction, or supply-chain activity.
Why datacenter expansion can raise emissions and water use
Construction and equipment add embodied emissions
A datacenter’s footprint begins before it serves a workload. Concrete, steel, and other construction materials carry emissions, as do the manufacture of semiconductors, servers, racks, and networking equipment. Futurism’s account identified construction and the hardware supply chain as important contributors to Microsoft’s emissions growth. These costs can be recorded in corporate emissions even though they are not emissions from electricity consumed inside an operating datacenter.
#1 Best Overall
- Save valuable floor space: 6U wall mount server cabinet Dimensions: 13.78" H x21.65" W x17.72" D.Maximum mounting depth is 14.2"
- Keep critical network equipment secure: glass door and side panels are lockable to prevent unauthorized access. Front door can be installed on either side of the front of the cabinet to satisfy your door swing orientation preference
- Easy equipment configuration: Fully adjustable mounting rails and numbered U positions, with square holes for easy equipment mounting with top and bottom punch-out panels for easy cable access
- Durability: Made of high quality cold rolled steel holds up to 110lb (50kg) (Easy Assembly Required)
- PCI & HIPPA and EIA/ECA-310-E compliant
Operating AI infrastructure requires power and cooling
AI clusters combine accelerators with networking, storage, backup power, and cooling. Electricity demand depends on the hardware, workload, utilization, facility design, and power supply. Cooling choices also matter: evaporative systems can consume water, while other designs may trade lower water use for higher electricity demand. The balance varies by site and system; a corporate annual total cannot identify which design or workload drove a change.
Water accounting needs more than one number
Water withdrawal is the amount taken from a source; water consumption is the portion not returned promptly to that source, often because it evaporates or is incorporated into a product. Datacenter cooling is one possible direct use, while electricity generation, semiconductor fabrication, and construction can add water impacts elsewhere in the supply chain. The reported 2022 and 2023 totals do not, on their own, locate use by watershed or establish how much was consumed rather than withdrawn.
What the figures say about AI—and what they do not
Microsoft’s AI and cloud expansion coincided with the reported rise, and the cited account links emissions growth substantially to datacenter construction and the materials and equipment needed to fill new facilities. That is a credible infrastructure explanation. It is not proof that generative-AI training or inference alone caused the increase: Microsoft’s totals cover a wider business, including cloud services and other computing activity, and the cited figures do not isolate AI.
Likewise, the totals do not reveal the share attributable to electricity versus embodied emissions, or whether growth was concentrated in particular facilities. A claim that every additional unit of water or carbon came from AI would go beyond what the available figures establish.
Rank #2
- Save valuable floor space: 12U wall mount server cabinet Dimensions: 24.25" H x21.65" W x17.72" D. MAXIMUM MOUNTING DEPTH is 14.2".
- Keep critical network equipment secure: glass door and side panels are lockable to prevent unauthorized access; Front door can be installed on either side of the front of the cabinet to satisfy your door swing orientation preference
- Easy equipment configuration: Fully adjustable mounting rails and numbered U positions, with square holes for easy equipment mounting with top and bottom punchout panels for easy cable access
- Durability: Made of high quality cold rolled steel holds up to 110lb (50kg) (Easy Assembly Required)
- PCI & HIPPA and EIA/ECA-310-E compliant
Microsoft’s climate target is a commitment, not an outcome
Microsoft’s stated goal is to become carbon negative by 2030. That is a corporate target, not evidence that emissions have already been neutralized or that the target will be met. The company has also described water replenishment and efforts involving renewable electricity, efficient cooling, lower-carbon construction, supply-chain decarbonization, and carbon removal.
Those actions address different parts of the problem and should not be treated as interchangeable. Reducing emissions at a facility differs from purchasing renewable electricity or compensating for emissions through carbon removals. Replenishing water elsewhere does not necessarily reduce consumption in the watershed where a datacenter operates. And an improvement in emissions per unit of computing or revenue can coexist with rising absolute emissions if total activity grows faster.
Is the increase temporary or structural?
Some construction-related emissions may be concentrated in a period of rapid expansion and could moderate after facilities are built. That possibility does not guarantee a later fall in total emissions. If demand keeps growing, Microsoft may continue adding datacenters, accelerators, grid connections, and cooling capacity; equipment manufacturing and replacement can also remain material.
Recommended Free Tools
The longer-term result depends on workload growth, hardware efficiency, utilization, facility location, power sources, cooling design, and the pace of supply-chain decarbonization. Efficiency can reduce resources per unit of computation, but total use may still rise when demand grows faster than efficiency improves.
Rank #3
- Sturdy:4u server rack is construct from cold rolled steel, with a weight capacity of 110lbs(50kg); Electrostatic powder coat prevents rust and corrosion,quality finish
- Direct use:Open and use, not having to assemble it.Network rack can be placed flat or mounted on the wall,also can be installed vertically under the table
- Design Features:maximum mounting depth of 14 in,cables can be fixed on the side panel;Open frame server rack achieves effortless inspection, replacement and assemble
- Installation:wall mount network rack is easy to install,with instructions or videos for reference;Equipped with multiple accessories, suitable for different needs
- Application:EIA/ECA-310-E Compliant;wall mounted 4u rack fits all 19" racks and cabinets to hold various IT, network, and AV equipment;wall mount rack available in 4U, 6U, and 8U to choose
Why corporate totals matter to local communities
Annual company-wide numbers cannot establish what a particular datacenter means for its neighbors. Local questions include whether a site draws from a water-stressed basin, how it affects grid capacity and transmission needs, what backup generators emit, and how land use, public incentives, noise, and employment compare. Water replenishment is most meaningful to local communities when its location, timing, watershed, and measured benefit are clear.
The cited coverage mentions community concerns in Arizona and Iowa but does not provide facility-level evidence sufficient to generalize those cases. A local assessment needs site-specific records and context; a corporate total is not a substitute.
What enterprise cloud customers and investors should ask
Microsoft’s headline figures are useful for spotting a trend, but they are not enough to compare workloads, facilities, or mitigation claims. Decision-makers evaluating cloud use or infrastructure plans should look for:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Absolute greenhouse-gas emissions alongside emissions intensity, with Scope 1, Scope 2, and Scope 3 figures distinguished.
- Water withdrawal and consumption reported separately, with facility or watershed context where available.
- Embodied emissions from construction and hardware, including lifecycle assumptions and equipment replacement.
- Electricity information that distinguishes market-based accounting from grid conditions and explains the timing of clean-power matching.
- Evidence that carbon removals are durable and disclosed separately from direct reductions.
- For water replenishment, the basin, timing, and verified benefit rather than a company-wide offset total alone.
For Azure customers seeking workload-level context, Microsoft provides an Emissions Impact Dashboard for Azure. Such cloud-specific estimates can inform decisions about cloud use, but they should not be mistaken for a complete accounting of hardware manufacturing, on-premises systems, or every corporate Scope 3 category.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
