Amazon, DOE, and Ames National Laboratory Launch AI-Led Materials Recovery Effort Targeting Graphite and Gallium Supply Chains

A new collaboration between Amazon, the U.S. Department of Energy’s Ames National Laboratory, and the Critical Materials Innovation Hub points to a more practical direction for domestic supply chain resilience: recovering critical materials from waste streams that are already embedded in the economy. The initiative brings together Amazon’s artificial intelligence capabilities and large-scale operational reach with the materials science expertise of Ames and the CMI Hub, with an initial focus on battery-grade graphite from discarded textiles and gallium recovery from end-of-life IT equipment.

The significance of the effort lies in how it links circularity to industrial strategy. In the United States, critical materials policy is increasingly shaped not only by sustainability concerns but also by supply security, import dependence, and the need to create viable domestic alternatives to foreign sourcing. This collaboration sits directly in that space, treating waste not as a disposal problem alone, but as a potentially strategic feedstock for next-generation industrial supply chains.

 

Waste Streams Are Being Reframed as Critical Materials Assets

 

The first workstream in the collaboration focuses on producing battery-grade graphite from post-consumer textiles, including discarded clothing. That is an important target because graphite remains a core battery material, and a domestic source derived from waste would carry both supply chain and circularity value. Instead of relying solely on newly mined or imported material, the project is exploring whether textile waste can be transformed into a useful industrial input with battery-grade quality.

This approach reflects a wider change in critical materials thinking. Traditionally, supply chain resilience has focused on extraction, refining, and geopolitics. Increasingly, however, resilience is also being linked to the ability to recover value from materials already circulating through consumer and industrial systems. If textile waste can be converted into graphite at viable quality and cost, it would expand the idea of where battery materials can come from and reduce pressure on more conventional supply pathways.

The work also aligns with the Department of Energy’s Genesis mission, which is aimed at securing critical materials from waste. That connection matters because it places the project within a broader federal effort to treat circular recovery as part of industrial and energy policy, not just environmental management.

 

Read more: Scientists Identify a More Efficient Route to Low-Energy Carbon Removal

 

Gallium Recovery From IT Hardware Targets a Less Visible Bottleneck

 

The second initiative centers on evaluating the technical and economic feasibility of recovering gallium and other critical minerals from end-of-life IT hardware. This is a strategically important direction because gallium is a relatively specialized material, yet it plays an important role in advanced electronics and other technologies. Recovering it from decommissioned hardware could help build a more distributed and circular source base, particularly if recovery methods can be made economically workable at scale.

This part of the collaboration also highlights the operational role Amazon Web Services is expected to play. AWS brings data infrastructure, physical AI expertise, and supply chain management capability, while CMI contributes technical depth in recovery science and access to its wider research network. The pairing is notable because materials recovery at scale is not only a chemistry challenge. It is also a logistics, sorting, identification, and systems-integration challenge. AI can become valuable here by helping identify recoverable value, optimize sorting and processing decisions, and improve the economics of what has often been treated as low-value end-of-life material.

That matters for the wider circular economy conversation. Many promising recovery pathways fail not because the science is impossible, but because the economics of collection, separation, and processing are weak. Using AI to improve those system-level decisions could make certain recovery pathways more commercially realistic.

 

AI Is Being Positioned as Industrial Infrastructure, Not Just a Digital Tool

 

A central theme in the announcement is that artificial intelligence is being used not as a generic innovation label, but as an operational tool within materials circularity. Amazon’s role suggests that AI can help bridge the gap between laboratory innovation and industrial deployment by improving how waste streams are identified, managed, and converted into usable materials.

This is important because circularity initiatives often struggle when they move from pilot science into real-world scale. A process may work in a laboratory but still fail commercially if feedstock quality varies too much, if logistics are inefficient, or if decision-making across the supply chain is too fragmented. AI can help address some of those issues by supporting faster classification, more precise process control, and better supply chain coordination.

In that sense, the collaboration reflects a more mature view of industrial AI. Rather than treating it as a standalone software capability, the project treats AI as part of the enabling infrastructure needed to make circular materials recovery more precise, more scalable, and more economically useful.

 

Explore OneStop ESG Marketplace: Waste management

 

Critical Materials Policy Is Expanding Beyond Mining

 

The broader policy significance of the initiative is that it reinforces a shift already underway in U.S. critical materials strategy. Domestic resilience is no longer being framed only in terms of opening new mines or expanding traditional refining capacity. It is also being framed around the ability to recover, upgrade, and reuse materials that already exist in waste streams across the economy.

That does not eliminate the importance of primary production, but it broadens the toolkit. Textile waste and retired IT hardware are not typically thought of as strategic resource bases, yet this collaboration suggests they may become part of a more diversified materials system if recovery technologies improve. For policymakers, this creates a stronger case for linking circular economy infrastructure with energy, industrial, and national security priorities.

The involvement of a company with Amazon’s scale also adds weight to the effort. It suggests that critical materials circularity is moving closer to mainstream industrial adoption rather than remaining confined to research programs and niche pilots. If large commercial actors begin to see waste-derived materials as strategically relevant, the market for recovery technologies could accelerate more quickly.

 

From Research Collaboration to Industrial Test Case

 

The partnership between Ames National Laboratory, the CMI Hub, and Amazon is still an early-stage collaboration, but it points to a larger opportunity. The real test will be whether the combined use of materials science, AI, and supply chain execution can turn difficult waste streams into dependable sources of battery and technology minerals.

If successful, the effort could help establish a more practical model for U.S. materials circularity, one in which discarded textiles and obsolete electronics are not treated as end points, but as inputs to a more resilient domestic industrial system. That would matter not only for sustainability goals, but for the broader question of how the United States reduces strategic exposure in the supply chains that underpin energy storage, electronics, and advanced manufacturing.