Lorem ipsum dolor sit amet, consectetur adipiscing elit. Curabitur eleifend tortor nec augue pretium
AI’s environmental impact – from its water consumption and carbon production to the inevitable rise in e-waste – is drawing an increasing amount of attention.
The situation is exacerbated by booming demand for new technology to meet AI’s component and hardware needs, putting pressure on IT manufacturers and leading to data-centre overprovisioning, as well as higher prices for relevant components.
From reducing manufacturing demand and building resilient supply chains to extending the usable lifespan of hardware, a circular IT economy would lessen AI’s effect on the environment.
According to a study published by Nature Computational Science, generative AI technology could create 1.2 to 5m tonnes of additional e-waste by 2030, taking the total to 82m tonnes. Just 22.3% of e-waste was recycled in 2022, and that figure is expected to fall to 20% by 2030.
At the same time, AI is putting pressure on the other end of the IT lifecycle: procurement. As demand grows, component prices are rising and new equipment lead times are lengthening. A recent survey of 280 data-centre professionals found that 83% don’t think supply chains can deliver the advanced cooling technologies required for AI workloads – and this is true for a range of components across the industry.
Furthermore, data centres are expected to consume 70% of the memory chips made available in 2026, causing industry shortages and raising prices as many scramble to meet their IT hardware needs.
When hardware’s lifespan is extended through circular practices (refurbishment, reuse, cascading assets and eventual recycling), manufacturing demand falls. In terms of IT procurement, this means a more resilient supply chain, as many turn to the refurbished market. Additionally, the embedded carbon in existing hardware is already spent; getting another three to five years from a device or component is almost always better for emissions than manufacturing a replacement.
The organisations that will thrive aren’t those chasing the new AI trend, but those with diversified supply chains that can adopt circular practices for a diversified market. Businesses could look to adopt a hybrid approach to AI – dealing with small, locally hosted models for some situations, and large language models provided by big companies for others. Exploring alternative supply chains, upgrading existing equipment and building low-cost sandbox systems, meanwhile, would allow companies to determine AI’s potential in a more circular manner.
Circularity also reduces the amount of hardware heading to landfill, resulting in less e-waste and more value reclamation from manufactured goods.
Expanding your pool of available resources means looking to the refurbished market for IT procurement. This market is mature and well stocked with leading and capable hardware, which is ready and waiting while lead times for new technology increase.
However, IT circularity can also mean exploring more sustainable IT disposal practices. Selling redundant IT hardware can maximise your overall budget, giving you the freedom to explore different options if your market is stretched thin. Recycling hardware should always be a last resort within the circular IT model, as it marks the end of a product’s life.
Circular IT habits not only reduce e-waste, carbon from manufacturing and the mining of virgin materials but also build a resilient supply chain for continued growth.
Charlotte Nixon is content and campaigns manager at Techbuyer Europe
