In a significant leap toward embracing a sustainable, circular economy, Amazon Web Services (AWS) has announced the successful integration of AI-driven robotics into its electronic device recycling processes. This pioneering initiative, made possible through a partnership with robotics startup Molg, aims to responsibly and efficiently reuse and recycle the immense volume of hardware decommissioned from AWS’s extensive global network of data centers.

Confronting the E-Waste Challenge of the AI Era

As demand for cloud computing and AI infrastructure grows, hyperscale data centers are expanding at breakneck pace. According to industry estimates, there are approximately 10,000 data centers worldwide, with Amazon among the sector’s largest operators. AWS runs hundreds of facilities globally, each packed with servers and networking equipment that routinely require upgrades or replacements within five to six years.

This rapid hardware turnover contributes to the world’s fastest-growing waste stream: electronic waste (e-waste). The United Nations estimates only 22% of used electronics are recycled, leaving the majority to accumulate in landfills, resulting in an annual loss of about $62 billion in material value. Many of the metals and materials in discarded electronics are not only valuable but also highly energy-intensive to produce, making effective recycling critical for both economic and environmental reasons.

AI and Robotics: The Molg Partnership

To address the mounting crisis of e-waste, AWS has teamed up with Molg, a startup founded in 2021 and currently employing 35 people in Northern Virginia. Amazon invested in Molg’s $10 million seed round via its $2 billion Climate Pledge Fund—a strategic move to drive innovations that shrink its carbon footprint and make operations more sustainable.

Molg’s machines use computer vision and AI to visually inspect decommissioned electronics, assess their condition, and precisely disassemble them for resale, reuse, or recycling. Compared to traditional, largely manual or destructive methods, Molg’s technology enables what CEO Rob Lawson-Shanks calls “de-manufacturing at the highest value.” This means recovering more components in usable condition, reducing resource loss, and boosting sustainability metrics.

Currently, Molg’s robots are operational at AWS’s re:Cycle Reverse Logistics facility in Pennsylvania. With operations escalating, Molg is producing “tens” of devices at a time, with ambitions to scale installations to additional AWS sites in Kentucky, Dublin, and Singapore.

Extending the Life of Data Center Electronics

“The useful life of these items is being extended by a year or two. We’re recovering more than we ever have previously. We’re reusing more of it,” remarked Nick Ellis, principal at the AWS Climate Pledge Fund. Automating the process not only yields environmental gains, but also generates cost savings and improves resource efficiency—core tenets of the circular economy model now being championed by Amazon and tech industry peers.

Amazon reported that in the previous year, 16% of data center components were moved back into use from its reuse inventory—a number AWS hopes to boost with Molg’s technology. “Having Molg robotics at our reverse logistics hubs is just a starting point,” added Manju Murugesan, AWS’s global circular economy lead. “It’s about figuring out how to bump up those figures and scale impact globally.”

In addition, Molg recently secured a $6.5 million grant from the U.S. Department of Energy to adapt their AI-powered disassembly technology for laptop recycling, aiming to further reduce the environmental impact across a broader device spectrum.

Economic and Environmental Benefits Drive Adoption

While specifics around cost savings remain confidential, AWS emphasizes the dual environmental and economic advantages of scaling AI-powered robotics for e-waste. By increasing component reuse and recovering valuable rare earth elements, the approach makes a measurable dent in the company’s sustainability goals.

To further address the challenge of rare earth material recovery, Amazon and Microsoft have also invested in Canadian startup Cyclic Materials, which specializes in extracting magnets and rare earth elements from recycled electronics. These rare materials are crucial for manufacturing new servers, consumer electronics, and renewable energy systems—resources often sourced via energy-intensive and polluting mining operations.

With the explosion of AI and cloud computing leading to increased demand for data center capacity, sustainability remains an industry-wide concern. Amazon, which pledged to achieve net zero carbon emissions by 2040, reported in July 2025 that total emissions actually rose by 6% over the past year, largely due to infrastructure expansions. Robotics and automation in device processing are regarded as pivotal in bending this curve back toward progress.

Setting a Standard for a Circular Tech Economy

The industry’s shift toward a “smarter” recycling paradigm isn’t limited to Amazon. Tech providers worldwide are exploring AI-driven systems to disassemble, sort, and recycle electronic waste more effectively, extending lifespans for hardware and shrinking the environmental cost per unit of compute.

“It’s about creating a circular economy that we’re all working so hard to achieve,” said Molg’s Lawson-Shanks. “Rather than the old ‘smash and grab’ approach—where reuse potential is often lost—we’re enabling components to be redeployed at their highest value.”

As robotics and machine learning systems continue to evolve, their deployment at the scale of AWS’s operations sets an industry standard for responsible electronic lifecycle management. With ongoing investments and expanded global operations, Amazon’s partnership with Molg could soon influence sustainability practices throughout the data center and technology ecosystems worldwide.

Related:
Recycling gets smarter: AI robots from Amazon-backed startup are sorting waste in Seattle
Amazon’s new AI system projects green and red markers onto packages to streamline deliveries