How the rapid growth of AI is creating a frightening increase in discarded technological hardware.
AI systems require powerful, specialized hardware — GPUs, processors, and servers — that become outdated quickly as technology advances. When newer, faster chips are released, older equipment gets replaced and discarded. The world generated 62 billion kg of e-waste in 2022, and only about 22% of it was properly collected and recycled (Geneva Environment Network, 2024). AI's demand for ever-faster hardware is accelerating this cycle.
E-waste can be toxic, is not biodegradable, and accumulates in the environment — in the soil, air, water, and living things. Open-air burning and acid baths used to recover valuable materials release toxic substances, exposing workers and communities to lead, mercury, cadmium, and arsenic, which can cause cancers, neurological damage, and miscarriages (Geneva Environment Network, 2024)
A circular economy approach — where devices are designed to be repaired, reused, and recycled — could dramatically reduce e-waste. Recycled metals are two to ten times more energy efficient than metals smelted from raw ore (Geneva Environment Network, 2024). Consumers can also extend device lifespans, and policymakers can mandate take-back programs and stricter recycling standards for tech companies.
All Topics
Water Consumption Electronic Waste Intensive Energy Usage Vital Resource Extraction Noise PollutionBourzac, K. (2024, November 4). Generative AI Has a Massive E-Waste Problem. IEEE Spectrum. https://spectrum.ieee.org/e-waste
Geneva Environment Network. (2024, October 9). The Growing Environmental Risks of E-Waste. Geneva Environment Network. https://www.genevaenvironmentnetwork.org/resources/updates/the-growing-environmental-risks-of-e-waste/