Amazon CEO Jassy Says AWS Capacity Sold Out for 2026, Mostly for 2027

Common ground

• The original 'sold out through 2027' headline was a misleading compression of a more hedged forecast, and the correction came too late to undo the damage.
• The $220 billion figure was never properly audited before publication, and when checked it didn't match AWS or total Amazon capex, possibly not even a single-year number.
• The 'sold out' claim is essentially unauditable because capacity bookings aren't disclosed with volumes, contract terms, or cancellation clauses.
• Both the CEO and the press have incentives that shaped the story: the CEO signals scarcity, and the press compresses for impact.
• The press has a special responsibility to verify and disclose uncertainty because it claims to be a check on power.

Points of contention

• Whether the Reuters self-correction shows the system working (neutral) or is just damage control after the harm was done (western).
• Whether the press's compression incentive is symmetric with the CEO's scarcity incentive (neutral says same gravity; western says asymmetric because only the press claims to be a check).
• Whether disclosure of a bias neutralizes it (western says it inoculates the claim; neutral refines that disclosure becomes an alibi).
• Whether the $220 billion audit dissolves the story (neutral) or is a separate failure that doesn't excuse the press's lack of discipline (western).

Blind spots

• Nobody addressed why this scarcity claim was amplified right when AI capex sentiment was wobbling—the timing may be strategic.
• The discussion treated 'the press' as a monolith, ignoring that the original X post and the Reuters desk are different actors.
• No one proposed a concrete standard for how journalists should handle unauditable executive claims beyond saying 'treat it as marketing.'
• The role of a16z (the original interview source) and its own incentives in framing the claim was never examined.

WorldAttention’s read

The roundtable agreed that the 'sold out through 2027' story was a case of a hedged forecast being compressed into a misleading headline, with a large capex figure that was never verified and a correction that arrived too late. The central claim about AWS capacity is structurally unauditable, so it should have been reported with explicit caveats rather than as fact. While the press self-corrected, the deeper failure was a lack of discipline in verification and transparency—especially given the press's role as a check on power. The CEO's scarcity signal and the press's compression incentive both shaped the narrative, and disclosure of bias does not neutralize its effect. The honest headline would have been: 'AWS says capacity is tight; here's why you can't check that.'