Amazon's quarterly report was not just a positive surprise, but a true manifesto on the efficiency of investments in artificial intelligence. The corporation not only beat Wall Street consensus forecasts on all key metrics, but also demonstrated that the AI race can yield dividends today, not in the distant future.
The company's shares closed the main session at $235.50, up 3.9%. However, after the release of financial results during after-hours trading, quotes soared another 8.85%, reaching $256.33. The market is clearly pricing in a new reality where Amazon acts not just as a retailer, but as a key beneficiary of the AI boom.
Numbers that speak for themselves
Net revenue for the second quarter reached $200.6 billion — up 20% from $167.7 billion a year earlier, while analysts on average expected around $197 billion. Operating profit jumped 43% to $27.5 billion, and operating margin expanded to 13.7% versus 11.4% last year. Market forecasts implied a margin closer to 12%, so here the company clearly outperformed expectations.
Diluted earnings per share deserve special attention: $5.75 versus the expected $1.82. However, comparing these figures directly would be incorrect. In the reporting period, Amazon recorded $53.4 billion in unrealized pre-tax income — mainly due to the revaluation of its stake in Anthropic, the developer of Claude models. But even excluding this one-time factor, the operating result remains extremely compelling.
AWS: the main growth driver
Revenue from the cloud division AWS reached $42.2 billion, corresponding to an annual run rate of $169 billion. The 36.7% increase was the highest for the segment since 2021. AWS operating profit grew 64% to $16.6 billion, and the segment's margin expanded to 39.4% from 32.9% a year earlier. Notably, the company did not have to resort to discounts to stimulate demand — growth was driven by organic interest in AI workloads.
In the context of competitors, this dynamic is especially telling. Microsoft previously reported Azure growth of 43%, while Meta's AI development costs led to margin compression from 43% to 31%. Amazon, by contrast, demonstrates that it is possible to scale AI capacity without sacrificing profitability. Moreover, agreements with Anthropic and OpenAI to use computing power based on its own Trainium chips strengthen vertical integration and reduce dependence on external suppliers.
The flip side of the coin
However, there is also a worrying signal. Free cash flow over the last 12 months turned negative: the company spent $7.6 billion instead of the $18.2 billion inflow a year earlier. The reason is capital expenditures: $54.2 billion in just one quarter and $169 billion over the year for building data centers and purchasing equipment. Operating cash flow, meanwhile, grew 33% to $161.4 billion, so the negative free cash flow is explained solely by massive construction, not by weakness in the core business.
Certain caution is also raised by forecasts for the third quarter: Amazon expects sales in the range of $197-202 billion, while analysts counted on approximately $204 billion. The company attributes the nearly 400 basis point difference to the shift in the Prime Day date. For investors doubting the returns on AI investments, it has now become harder to challenge their viability.
My view: The key question is not whether Amazon's AI spending pays off — the report has already given an unequivocal answer. What matters more is whether AWS can sustain 37% growth when the comparison base becomes significantly higher next year. If the pace holds, Amazon could redefine the rules of the game in the cloud industry, leaving competitors behind.