The Cloud Sustainability Shift: Can AI Help Green IT?

Data centres power the cloud, yet their energy use is enormous. As AI and cloud workloads grow, businesses must ask: can AI itself be part of the solution in making cloud computing sustainable?

In this article we look at how green cloud strategies, carbon-aware scheduling, and intelligent optimisation are reshaping sustainable cloud computing in 2025 and beyond.

What Is the Energy Footprint of Cloud and AI?

Cloud infrastructure relies on large data centres that draw power, cooling, and water. With ever-rising demand from AI models, the footprint expands.

According to CBRE, data centres consumed 415–460 TWh globally in 2024, about 2% of world electricity usage, and AI workloads may drive that much higher. Data Centre Magazine The UK has over 500 data centres, making it one of the top hosts globally. Energy UK+1

If unchecked, energy costs, regulatory pressure, and public scrutiny will force a shift.

Why Use AI to Make Cloud More Sustainable?

Because AI can monitor, predict, and optimise operations at scale. It can identify inefficiencies that humans miss, then adjust workloads or cooling systems to match.

CBRE highlights three strategies powered by AI:

• Adjust cooling and power consumption based on real-time data and weather forecasts

• Integrate with smart grids to use cleaner energy when available

• Reuse waste heat or coordinate load across facilities to reduce peaks Data Centre Magazine

In short, AI gives you the tools to reduce carbon output while maintaining performance.

How Does Carbon-Aware Scheduling Work?

Carbon-aware scheduling is a strategy where compute tasks are timed or shifted to take advantage of lower carbon electricity periods or regions.

How it works:

• Monitor grid carbon intensity in real time

• Delay non-urgent workloads to periods of lower emissions

• Move jobs between regions where power is cleaner

• AI models forecast when and where to run tasks for minimal carbon cost

This allows organisations to reduce emissions without compromising service.

One research system called HUNTER used AI techniques (graph networks) to schedule tasks across energy, thermal and cooling models. Their tests showed energy reductions up to 12%, cost down 54%, fewer temperature violations and better resource use. arXiv

What Should Businesses Do to Adopt Green Cloud?

1. Choose sustainable cloud providers: Opt for providers that publish carbon metrics, use renewable energy sources, or are signatories to pledges like the Climate Neutral Data Centre Pact. Wikipedia

2. Apply carbon-aware policies: Work with your cloud or AI provider to schedule workloads based on grid emissions or cost signals.

3. Use AI monitoring and optimisation tools: Tools that monitor data centre power, temperature, utilisation and adapt dynamically.

4. Revisit heavy workloads quarterly: Assess what is being run, what can be shifted or consolidated. This supports freshness updates: revisiting high-traffic posts and processes to keep energy use in check.

5. Demand transparency: Ask providers for published sustainability metrics, energy sourcing, cooling techniques and audit compliance.

AI and Green IT at a Glance

Area	Common Cloud Challenge	AI-Driven Green Approach
Cooling	Excess energy in heat removal	Predictive cooling adjustments
Scheduling	Jobs run at peak grid emissions	Carbon-aware scheduling
Workload placement	Static assignment regardless of energy source	Dynamically shifting tasks
Provider choice	Providers opaque about energy use	Prefer providers with clear sustainability reports

Final Thoughts

Cloud computing and AI can feel at odds with sustainability. Yet, with intelligent design and strategy, AI becomes a tool to reduce carbon, not inflate it. The shift to “green cloud” is underway. For UK businesses, aligning AI adoption with sustainable cloud practices will matter more every year.