The Basel Action Network now projects that between 395 million and 617 million metric tons of electronic equipment could be retired between 2025 and 2050 in connection with artificial intelligence — triple the current volume of equipment in service. The revised figures appear in the first installment of BAN's four-part white paper series, "The Coming AI Waste Wave," and they land 28% above the NGO's previous estimates for the stream.

The material BAN counts goes well beyond the server fleet. The report breaks down a data center's electro-mechanical infrastructure by mass: cooling systems account for 35%, power supply and distribution for 34%, backup power systems for 15%, servers, accelerators and racks for 13%, and networking equipment for 3%. Processors are only the visible fraction of the tonnage.

A 'contagion' outside the data hall

The largest upward revision comes from what BAN calls "AI waste contagion" — electronics outside data centers retired early as personal devices and telecommunications networks need upgrades to work with AI. That obsolescence channel could add between 16.2 million and 30.7 million metric tons of retired equipment annually to the global total by 2050.

"This AI contagion obsolescence as we call it, will not take place in data centers but with our own IT equipment in our homes, our businesses, and our gaming rooms," the report states.

Stacked on top of that, BAN projects total global e-scrap generation could reach up to 211 million metric tons annually by 2050 — triple today's annual total. For comparison, the UN Global E-Waste Monitor estimated 62 million metric tons of e-scrap generated in 2022, with only 22.3% properly recycled. Scaled on that trajectory without AI-driven demand, 2050 generation would sit near 165 million metric tons.

The model's inputs include the pace of capacity growth, the equipment volume each added gigawatt requires, and how long that equipment lasts. BAN flags scenarios that could push tonnage higher still: data sovereignty laws mandating redundancy, data centers becoming critical infrastructure and conflict targets, a wholesale replacement of aged telecom networks, and a "rip-and-replace" transition to superconducting power systems, liquid cooling and new battery chemistries that strands otherwise functional infrastructure.

The report also concedes downside risk to its own numbers. Equipment may outlast projected lifespans, servers could be modularly upgraded, a secondary market for reuse and refurbishment may emerge, and software efficiency gains could cut hardware demand. AI and data center demand could also plateau before reaching the levels BAN modeled through 2050.

BAN's stated objective: slow the buildout

Jim Puckett, BAN's founder, was blunt about the series' purpose: slowing the AI boom, not stopping it.

"I earnestly hope we can slow the horserace, the freight train, the juggernaut that is the 'AI revolution' down," Puckett said by email. "From what I have learned in just the last few months, and only looking at this one segment of its impact — toxic e-waste — I am scared for that outcome, while realizing there are so many other results far more harrowing."

Puckett acknowledged the irony that he used Anthropic's Claude model to help draft the white paper, citing beneficial applications such as cancer detection and early warning systems. "Given the speed at which things are moving with the AI horserace, I felt I could not afford to worry about this moment's ironies in order to match that sense of urgency," he said, adding that he is "harnessing the power of AI in the most positive way now in the hope of being able to rein in its likelihood of it creating havoc."

He drew a parallel to AI developers themselves: "It's the same irony even the masters of AI are facing down when they remarkably beg governments to control their product."

What comes next

Three further installments of "The Coming AI Waste Wave" are planned, covering whether reuse, refurbishment and repurposing can shrink the stream, the toxicity profile of AI-driven e-scrap, and the policy responses BAN recommends. With documented reuse and recycling rates stuck at 22.3% globally, the policy installment will determine whether regulators treat data center decommissioning as a controlled material stream or leave it to voluntary industry practice.