Jassy soothes investor worries over Amazon’s $220bn AI spending with AWS growth case

Amazon CEO Andy Jassy on Thursday assuaged investor concerns about the company’s escalated artificial intelligence buildout by pairing a higher 2026 capital-expenditure target with stronger-than-expected cloud results and a clear pitch that the spending will generate returns.

Jassy told investors “At this level of spend and higher, we have clear line-of-sight to strong financial returns,” and argued that “as we get a few years out and the revenue growth outpaces the incremental capex growth… the resulting revenue, free cash flow and return on invested capital is very compelling,” remarks cited by CNBC from the company’s quarterly update show.

Why this matters: investors had fixated on Amazon’s jump in planned spending, but the market recalibrated after Amazon reported a surge in AWS demand that, for now, makes the capex appear monetizable rather than purely speculative.

2026 capex rises to about $220 billion; investors feared the bill

Amazon boosted its 2026 capital-expenditure outlook to roughly $220 billion from an earlier $200 billion target, a move that initially stoked investor anxiety about a multiyear AI cash burn. The company’s Q2 capex alone ran near $53.1 billion, up from about $44 billion in Q1, underscoring the stepped-up pace of data‑center and hardware investment.

Markets reacted quickly: Reuters reports Amazon shares jumped more than 12% in premarket trading after investors shifted attention to the cloud numbers rather than the headline spending figure. Still, analysts and some investors had worried that such a large and sustained outlay — including a recent bond issuance — could compress returns if revenue failed to follow.

AWS AI run rate tops $25 billion and triples year over year

Amazon’s argument rests on AWS traction. The company said its AI-related cloud business runs at more than $25 billion and posted triple‑digit year‑over‑year growth, an unusually fast expansion for a business of that scale, according to CNBC’s earnings coverage. Jassy highlighted that existing customer demand and the company’s vertically integrated stack — from custom chips to its own data centres — give Amazon “line‑of‑sight” to converting capex into revenue and free cash flow.

That pitch contrasts with another Big Tech story: Reuters flagged Alphabet as a peer under scrutiny for AI‑related cash burn, where investors have been more skeptical about the pace of monetisation. Amazon’s counter is practical: show growing, high‑margin cloud revenue today to justify heavy infrastructure spending for tomorrow.

Critics remain. Reuters noted the market initially focused on the spending increase and questioned whether multibillion‑dollar AI bets will ultimately justify the outlays. Some investors liken the move to a high‑stakes capacity build — useful if demand materialises, painful if it does not — and flagged the risk that capex could be front‑loaded before revenue catches up.

CNBC has repeatedly emphasised Jassy’s framing that Amazon is not making the investment “on a hunch,” pointing to concrete AWS demand and partnerships, including work with Anthropic, as evidence the company’s AI strategy ties to paying customers rather than pure R&D experimentation.

A skeptical sell‑side voice told Reuters that markets will test the thesis: sustained secular demand, not one quarter of strong growth, must validate a $220 billion capex plan.

Amazon’s approach also reveals a strategic divergence inside the hyperscaler cohort. Where Alphabet and others face questions about when AI spending will pay back, Amazon is leaning on an already‑monetising cloud division and its own silicon to shorten that timeline — a bet that demands both continued AWS growth and discipline on future capital allocation.

Jassy’s comments and the quarter’s figures briefly turned investor attention from the headline capex number to revenue trajectory; the test now is execution. The next concrete metric to watch will be AWS revenue growth and margin trends in Q3 and whether free‑cash‑flow improvement begins to outpace the cumulative capex build, validating Jassy’s “line‑of‑sight” claim.