🌱 How do generative AI systems create e-waste?
Generative AI algorithms and models — which include ChatGPT, large language models, and other “types of AI that generate texts, images, videos or music from massive datasets” — are “dependent on rapid improvements in hardware infrastructure and chip technologies”. The same holds true for the “computer resources required for training and using AI in data centers”. E-waste is produced when hardware becomes outdated and is consequently updated or replaced.
🌱 How much e-waste is estimated?
Overall, “[t]he increasing popularity of generative AI is projected to result in the rapid growth of e-waste”. A study published in Nature Computational Science in 2024 estimates that the e-waste “created from generative AI data centers could rise to 5 million tons per year by 2030”. Therewith, the researchers are also “predicting a thousand-fold increase in e-waste from AI computer servers by 2030”.
🌱 Why might even more e-waste be produced?
As the market for AI is developing rapidly, the estimates from the 2024 study are “potentially on the low side”. Amongst others, “[f]actors such as geopolitical restrictions on semiconductor imports and rapid server turnover” could further increase the e-waste created through generative AI. It is also important to note that other forms of AI also create e-waste, and these too are expected to create their own e-waste.
🌱 How could the e-waste from AI be reduced?
Unsurprisingly, it is “easier and more cost-effective to address the e-waste challenges posed by AI now" rather than later on. Overall, prolonging the use of hardware, reusing components, and extracting valuable materials through recycling could significantly reduce e-waste. The 2024 study estimates that these strategies “could reduce e-waste creation by 16% to 86%”. This is most likely to succeed “if supported by policies” and “widely implemented across industries and regions".
Read more about e-waste from generative AI here:
- https://www.dw.com/en/e-waste-from-ai-computers-could-escalate-beyond-control/a-70619724
- https://www.nature.com/articles/s43588-024-00712-6
- https://spectrum.ieee.org/e-waste
- https://hbr.org/2025/02/ais-growing-waste-problem-and-how-to-solve-it