AI is generating significant electronic waste, a critical issue that remains largely unnoticed.
A new report highlights that e-waste from the artificial intelligence boom has been significantly underestimated. The study predicts that by 2050, this waste could accumulate to a volume sufficient to fill 23 million shipping containers, which would stretch approximately six times around the world if lined up end-to-end.
This estimate is substantially higher than previous studies because the authors, from the nonprofit Basel Action Network (BAN), incorporated all infrastructure required to support data center servers. This broader perspective offers a more accurate assessment of the environmental toll left by AI technology.
"AI may feel weightless, but every model depends on an enormous amount of highly specialized, cutting edge hardware," states Jim Puckett, founder and chief of strategic direction at BAN. "If companies and governments do not begin planning for this new waste tsunami, today’s AI buildout could become an even more cataclysmic toxic waste crisis than we are already experiencing."
Currently, less than a quarter of the 68.3 million tons of e-waste produced globally each year is formally collected and recycled. Instead, the majority ends up in informal disposal sites where burning or burying equipment exposes workers and local environments to hazardous materials like lead and chromium.
The United States, which possesses more data centers than any other nation, has not ratified the Basel Convention aimed at restricting international hazardous waste trade. Investigations indicate that US recyclers continue to ship e-waste abroad, often entering "backyard recycling" operations. Millions of children living near these informal sites face severe health risks from toxic exposure, according to the World Health Organization.
BAN anticipates that global e-waste generation will triple by 2050, reaching up to 211 million metric tons annually. The organization attributes 15 to 20 percent of this garbage to AI. Unlike earlier estimates that focused solely on servers and GPUs, BAN has expanded its count to include power supplies, cooling systems, backup power, networking equipment, and what it terms "AI Waste Contagion." This latter category encompasses telecommunications infrastructure and personal devices that may become obsolete faster due to rapid AI advancements.
BAN projects that the retirement of AI-related equipment will generate 70,000 metric tons of waste for every gigawatt of data center capacity. Citing a McKinsey forecast that total data center capacity could reach 219 gigawatts by 2030, the organization calculates that e-waste from AI equipment retired between 2025 and 2050 will amount to between 395 and 617 million metric tons.
This translates to an annual retirement of between 8.6 million and 13.1 million metric tons of equipment driven by AI. Previous estimates were more conservative, often focusing exclusively on servers and accelerators. However, BAN notes that these earlier approaches failed to account for approximately 87 percent of a data center's electro-mechanical infrastructure.
A separate study from 2024 predicted e-waste from AI could reach 1.2 to 5 million tons by 2030. Another analysis released in February suggested that AI servers could generate between 131,000 and 225,000 tons of waste annually by the end of the decade—a volume comparable to the total annual e-waste production of Denmark.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.