Amazon's AI Bet: Record Cloud Orders and an Autonomous Shopping Agent

Amazon invests record sums in AWS AI infrastructure and launches 'Alexa for Shopping' with autonomous purchasing, as cloud revenue surges 28% and e-commerce profit jumps 47%.

Amazon is extending its artificial intelligence push on two parallel fronts, ploughing unprecedented sums into cloud infrastructure while reengineering the consumer shopping experience with a fully autonomous purchasing agent. The strategy is already reflected in a growing cloud backlog that stands at $364 billion, a figure that underscores how deeply enterprise demand for AI computing has penetrated the company's AWS division.

Cloud backlog swells as chip commitments soar

AWS generated $37.5 billion in revenue during the first quarter of 2026, a 28% year-over-year surge that cemented its role as the company's primary profit engine. The cloud unit now supplies nearly 60% of Amazon's total operating income, powered largely by demand for specialised AI hardware. The company has locked in firm commitments worth $225 billion for its in-house Trainium chips, and a single partnership with AI startup Anthropic accounts for $100 billion of the total order book.

To meet that demand, Amazon is earmarking roughly $200 billion in capital expenditure this year — a 50% increase from 2025. The bulk of that spending will go toward new data centres and expanded chip production capacity, a scale of investment that few peers can match.

E-commerce fundamentals surprise

Even as the cloud business dominates headlines, the traditional retail operation is delivering its own record. Operating profit from e-commerce climbed 47% in the first quarter to $9.7 billion, far outpacing revenue growth. Total company revenue reached $181.5 billion, comfortably above the $177.3 billion analysts had anticipated.

The earnings surprise was even more pronounced on a per-share basis: adjusted EPS came in at $2.78, against a consensus forecast of just $1.63. That beats gives management ample room to execute the strategic overhaul underway in the retail division.

Consumer AI takes a new form

The most visible sign of that overhaul is the replacement of Amazon's AI shopping assistant, Rufus, with a new system called "Alexa for Shopping". Unlike its predecessor — a generative chatbot introduced in 2024 — the new agent is designed to act, not just answer. Its "Buy for Me" feature can complete purchases on external merchant sites, effectively giving Amazon a checkout layer across the wider web.

That shift turns Alexa into a universal shopping interface capable of tracking prices, scheduling recurring orders, and capturing purchase intent even when the sale happens outside Amazon's own marketplace. The company is betting that agentic software, baked into its massive installed base of Alexa devices, will tighten its grip on the online retail ecosystem.

Stock near highs despite job-cut confusion

Investors have cheered the dual narrative. Amazon shares closed at €230.15 on Thursday, up 9.4% over the past month and 19% since the start of the year — a pace that puts the stock within striking distance of its all-time high.

That rally has persisted despite a brief mid-May dip after the earnings beat, when the market appeared to focus on operational noise. Reports of a potential 14,000-job reduction briefly unsettled traders, but a company spokesperson dismissed the figure as inaccurate and not supported by facts. Amazon did announce a real cut in its seller-services unit in May, as it redirects headcount from traditional roles toward AI engineering. The company plans to hire 11,000 new software developers this year, part of a broader rebalancing toward machine-learning talent.

Valuation still offers a discount

At a trailing price-to-earnings ratio of 32, the equity trades well below its 10-year average of 46, a gap that several analysts view as a buying opportunity. Consensus price targets hover around $312 per share, implying roughly 30% upside from current levels.

The next near-term catalyst arrives in June with Prime Day, which is expected to generate record sales in the U.S. Amazon is also testing 30-minute delivery in selected metro areas, an operational push that aligns with the new agentic shopping layer. If the consumer AI bet sticks, the company could lock in a level of purchase-data visibility that even its largest retail rivals will find hard to match.