For Immediate Release

AI E-Waste May be 40 to 60 Times What Has Been Estimated, New Analysis Finds

First-of-its-kind full-infrastructure analysis by Basel Action Network projects 395 to 617 million tonnes of AI-driven electronic equipment could be retired between 2025 and 2050. If packed into shipping containers and placed end to end, they would circle the Earth about six times.

SEATTLE, WA. September 16, 2026. Today, the Basel Action Network (BAN) released Part 1 of The Coming AI Waste Wave, a new four-part white paper series examining an overlooked environmental consequence of the global artificial intelligence boom: the vast amount of electronic equipment that will become e-waste as AI infrastructure expands dramatically and rapidly becomes obsolete.

BAN's new analysis, entitled "How Big is the AI Waste Wave?” finds that by 2030, an estimated 8.6 to 13.1 million tonnes of AI-driven electronic equipment could be retired each year - roughly 40 to 60 times the most widely cited academic projection. The difference is primarily one of scope: previous quantitative studies have focused on servers and graphics processing units (GPUs), while BAN examines the full electronic infrastructure supporting AI data centers and the additional hardware displaced as AI adoption accelerates.

McKinsey estimates nearly $7 trillion in capital expenditure between 2025 and 2030 to build the infrastructure required to power AI. BAN's model starts with what the industry says it intends to build and calculates the potential waste implied by that expansion. Its calculation includes networking equipment, power distribution, data storage and backup systems, and cooling infrastructure, as well as servers and computing equipment -- virtually all of which qualifies as electronic waste by definition. BAN also adds what it calls the AI Waste Contagion: the electronic equipment outside data centers that may be retired early as AI spreads across personal computers, phones, edge devices and telecommunications networks making traditional electronic equipment obsolete. 

BAN also signals the grave concern that much of the new AI waste is likely to be contaminated with PFAs, the so-called "Forever Chemicals" which regulators are rapidly seeking to ban due to their persistence and toxicity in the human body and in the environment. This subject will be explored in depth in Part 3 of the white paper series.   

"AI may feel weightless, but every model depends on an enormous amount of highly specialized, cutting edge hardware," said Jim Puckett, Founder and Chief of Strategic Direction at the Basel Action Network. "To date the environmental debate around AI has focused on electricity, carbon and water while largely overlooking what happens to the hardware itself. 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." 

The report's key findings include:

• AI’s infrastructure build-out carries a staggering price tag: Industry insiders project an expenditure of $7 trillion from 2025 to 2030 just to begin the build-out of the necessary infrastructure to power AI. $7 trillion is roughly the cost of ending world hunger for the next 75 years being spent in five. Or, the sum of money could instead fund a $10,000 higher-education trust for every child born on Earth (700 million children) during those 5 years.
  
• Most of AI's data-center infrastructure is absent from previous estimates: Servers, accelerators and racks account for an estimated 13% of a data center's electro-mechanical infrastructure by mass. BAN's model incorporates the remaining 87%, including networking, power distribution, storage and backup, and cooling systems.
  
• Annual AI-driven equipment retirement could rise sharply: BAN projects that 8.6 to 13.1 million tonnes of electronic equipment could be retired annually by 2030—approximately 40 to 60 times the most widely cited academic projection. By 2050, the annual amount could reach 31 to 46 million tonnes.
  
• The cumulative physical footprint could be immense: Between 2025 and 2050, BAN projects that 395 to 617 million tonnes of AI-driven electronic equipment could be retired. That volume would fill approximately 15 to 23 million 40-foot shipping containers - enough to circle the Earth about six times if placed end to end.
  
• AI's waste footprint extends beyond data centers: BAN estimates that the AI Waste Contagion could add 16.2 to 30.7 million tonnes of retired equipment annually by 2050. Under both scenarios, this downstream retirement eventually exceeds the 15.5 million tonnes per year projected from inside data centers.
  
• AI could substantially increase the global e-waste burden: When added to the conventional e-waste trajectory, total global generation of electronic waste could reach 196 to 211 million tonnes per year - 19% to 28% above BAN's extrapolation of the UN baseline and more than triple today's annual total.

The paper also warns of the declining life-span of AI equipment generally as well as three potential "rip-and-replace" transitions that will occur when the data centers switch to superconducting power systems, liquid cooling and new battery chemistries. These projected innovations could retire functional infrastructure even earlier than the BAN model's scheduled replacement cycles assume.

Today, at our current rates of e-waste generation, the world manages only a fifth of its e-waste responsibly, according to the Global e-Waste Monitor. BAN’s own investigations (see Brokers of Shame) have found that much of this waste is not properly recycled but exported to highly polluting operations in developing countries, often by companies certified as responsible recyclers, where it ends up harming workers, communities and the local environment. Against that backdrop, BAN warns that an increased wave of AI-driven e-waste on the scale it projects would further overwhelm a waste management system already failing to manage the world's e-waste.

"We cannot ethically manage the e-waste we produce currently”, Puckett said. "Technology companies must take steps now to develop a budget, new infrastructure and a plan to mitigate the AI toxic waste wave. The time to prevent toxic waste is before it is created, not 20 years from now, after hundreds of millions of tonnes of equipment are pushed out of the thousands of data centers across the globe to vulnerable communities.”

END

For more information:

Jim Puckett

Founder and Chief of Strategic Direction

Basel Action Network (United States)

email: jpuckett@ban.org

White paper

You can access the whitepaper here: https://wiki.ban.org/images/1/16/AI_Waste_Wave_Part_1.pdf

How Big Is the AI Waste Wave? is Part 1 of BAN's four-part whitepaper series entitled, The Coming AI Waste Wave. Subsequent releases will examine whether it will be possible to reuse, refurbish and repurpose to reduce its impact (Part 2); the toxicity of this emerging waste stream (Part 3); and finally the policy responses needed to manage this emerging crisis responsibly (Part 4). 

Note on Methodology: BAN developed a bottom-up model based on global installed data-center capacity, projected industry growth, the equipment required per gigawatt and category-specific replacement cycles. The analysis accounts for five categories of data-center equipment, as well as AI-driven equipment retirement outside data centers, and presents conservative and aggressive scenarios through 2050. The full methodology, sources, assumptions and sensitivity analysis are included in the whitepaper’s appendices.

About Basel Action Network

Founded in 1997, the Basel Action Network is a 501(c)3 charitable organization of the United States, based in Seattle, WA. BAN is the world's only organization focused on confronting the global environmental justice and economic inefficiency of toxic trade and its devastating impacts. Today, BAN serves as the information clearinghouse on the subject of waste trade for journalists, academics, and the general public. Through its investigations, BAN uncovered the tragedy of hazardous electronic and toxic waste dumping in developing countries. For more information, see https://www.ban.org/