Amazon's Copper Bet: Securing the AI Infrastructure S-Curve

The core investment thesis for Amazon is clear: it is building the fundamental rails for the next technological paradigm. This is not about incremental growth; it is about positioning for exponential adoption. At the heart of this bet is Amazon Web Services, which is now accelerating on the steep part of the S-curve.

AWS's growth rate is a key signal. In the third quarter of 2025, the unit posted revenue growth of 20.2 percent year-over-year, its fastest pace since 2022. More importantly, this surge is happening on a massive base, with AWS now operating at a $132 billion annualized run rate. As CEO Andy Jassy noted, this is a fundamentally different dynamic than competitors' higher percentage gains on much smaller revenue bases. The sheer scale of this growth validates AWS's dominant infrastructure position.

This growth is being fueled by an unprecedented build-out of physical capacity. Over the last 12 months, AWS has added 3.8 gigawatts of data center power capacity. To grasp the magnitude, that figure represents double the power capacity AWS had in 2022. The company is on track to double again by 2027, a cadence that mirrors the exponential demand curve for AI compute. This isn't just expansion; it's a strategic pre-positioning to capture the next wave of adoption.

The most concrete evidence of this infrastructure bet is the company's commitment to the U.S. government. Amazon announced an investment of up to $50 billion to build new data centers, adding nearly 1.3 gigawatts of AI and supercomputing capacity. Construction is set to begin in 2026. This project is a direct play on the government's own AI ambitions and secures a massive, long-term demand stream for AWS's infrastructure. It underscores that Amazon is not just building for today's cloud workloads, but for the next decade's compute-intensive applications.

Securing the Input: The Copper Supply Chain Bet

Amazon's bet on AI infrastructure now extends beyond data centers to the very raw materials that build them. The company has locked in a critical input by signing a two-year agreement with Rio Tinto to source copper from a new Arizona mine using bioleaching technology. The deal covers approximately 14,000 metric tons of copper cathode over four years, with a further 16,000 tonnes from conventional leaching, totaling nearly 30,000 tonnes.

This move is a direct response to a looming supply constraint. AI-driven copper demand is projected to surge 50% by 2040, yet analysts warn supplies could fall far short. By securing a domestic source, Amazon is mitigating a key vulnerability in its exponential growth trajectory. Copper is essential for wiring, circuit boards, and heat sinks in every data center, with a single large facility requiring tens of thousands of tons. The Johnson Camp mine, using Rio Tinto's Nuton technology, is one of the first new U.S. copper production sources in over a decade, providing a strategic foothold in a constrained market.

The strategic importance goes beyond simple supply. Amazon is targeting lower-carbon materials across its supply chain, and this collaboration aligns with that goal. The Nuton process, which uses bacteria to extract copper and produces 99.99% pure copper cathode at the mine gate with lower water use and emissions, offers a more sustainable input. In practice, this means AWS can build its infrastructure with a more resilient and environmentally aligned supply chain, a subtle but material advantage as ESG pressures mount.

Yet the scale of the commitment reveals the magnitude of the challenge. As noted, this supply represents only a sliver of Amazon's needs, insufficient for even one major data center. This is not a complete solution, but a first-mover hedge. It secures a portion of future demand and provides valuable data on integrating new, lower-carbon copper sources into massive industrial production. For a company building the rails for the AI paradigm, controlling the input is as crucial as controlling the output.

Financial Impact and Valuation: Growth vs. Cost

The strategic bets on AI infrastructure are now translating directly into Amazon's financials, with capital expenditure serving as the clearest indicator of commitment. The company has raised its forecast for 2025 capital spending to $125 billion, up from an earlier estimate of $118 billion. This massive outlay is the fuel for the exponential growth engine, funding the 3.8 gigawatts of data center capacity added over the last year and the planned $50 billion government project. The market is recognizing this acceleration, with the stock surging 11.55% in the week following the Q3 2025 earnings that highlighted AWS's 20.2% revenue growth and its $200 billion backlog. This move signals that investors are pricing in the steep part of the S-curve.

Valuation presents a nuanced picture. The stock trades at a forward P/E of about 36, which is not cheap. Yet the PEG ratio, which adjusts for growth, sits at 0.65. This suggests the market is assigning a premium for growth but with a margin of safety relative to the earnings expansion implied by the capital build-out. In other words, the price may already reflect the massive infrastructure investment, leaving room for the exponential adoption of AI workloads to drive earnings higher than currently priced.

The bottom line is that Amazon is investing heavily to secure its position on the next paradigm's infrastructure layer. The raised capex forecast is the cost of admission to the AI race, while the PEG ratio indicates the market sees the growth payoff ahead. The recent price surge shows recognition of the acceleration, but the real test will be whether the company can convert this massive capital investment into sustained, high-margin revenue as the adoption curve steepens. For now, the financials align with the deep tech thesis: building the rails for exponential growth.

Catalysts and Risks: The Path to Exponential Adoption

The trajectory of Amazon's AI infrastructure bet now hinges on a few critical future events. The primary catalyst is the successful execution of the $50 billion government data center build-out, with construction set to begin in 2026. This project is a direct, multi-year demand anchor for AWS's exponential capacity expansion. Its completion will validate the company's massive capital investment and secure a long-term revenue stream from a strategic customer. The watchpoint here is not just the start of construction, but the steady ramp of the 1.3 gigawatts of AI and supercomputing capacity, which must be delivered on schedule to meet the government's own AI ambitions.

The key risk, however, is a persistent supply constraint that could pressure the very infrastructure Amazon is building. While the copper deal with Rio Tinto is a smart hedge, it is a sliver of the total need. As industry watchers warn, only 70% of 2035 demand could be met. If broader copper supplies fall short, it could lead to higher costs and potential delays for AWS's expansion. This isn't a distant theoretical risk; it's a tangible friction point that could slow the build-out pace and erode margins on the massive capex already committed.

Finally, the competitive landscape demands constant vigilance. AWS's recent 20.2 percent year-over-year revenue growth is impressive, but it's on a $132 billion annualized run rate. The real test is maintaining that 20%+ growth on an even larger base as rivals like Microsoft and Google scale aggressively. Their own cloud revenue growth rates are in the high teens to low twenties, showing the competition is not standing still. AWS's ability to keep accelerating, fueled by its backlog and new projects, will determine whether it maintains its lead on the steep part of the S-curve or gets caught in a costly, slower-growth plateau.