AI data center e-waste could reach 617 million metric tons by 2050, report finds

A new report from the nonprofit Basel Action Network (BAN) warns that e-waste from the AI boom has been vastly underestimated and could reach between 395 and 617 million metric tons by 2050. The estimate is significantly higher than previous studies because it accounts for the full range of infrastructure supporting AI data centers, including servers, GPUs, power supply and distribution systems, cooling systems, backup power, and networking equipment.

BAN’s report also introduces the concept of "AI Waste Contagion," a category that encompasses telecommunications infrastructure and personal devices likely to become obsolete earlier due to AI advancements. The organization projects that global e-waste will triple to up to 211 million metric tons annually by 2050, with AI responsible for 15 to 20 percent of that total.

The report estimates that 70,000 metric tons of e-waste will be generated per gigawatt of data center capacity. It cites a McKinsey projection that total data center capacity could reach up to 219 gigawatts by 2030, implying a substantial increase in AI-related electronic waste in the coming decades.

Less than a quarter of the 68.3 million tons of e-waste produced globally each year is formally collected and recycled, according to BAN. Most e-waste ends up in informal waste streams, where burning or burying equipment can expose workers and local environments to toxic materials such as lead and chromium. The U.S., which hosts more data centers than any other country, has yet to ratify the Basel Convention, which aims to limit the international trade of hazardous wastes.

BAN’s findings contrast with earlier studies that focused narrowly on servers and accelerators. By broadening the scope, the report argues that past estimates missed about 87 percent of a data center’s electro-mechanical infrastructure. For example, a 2024 study predicted AI-related e-waste could reach between 1.2 million and 5 million tons by 2030, while a February 2024 study estimated AI servers alone could generate between 131,000 and 225,000 tons of e-waste annually by the end of the decade.