Oracle and Adobe face Thursday’s AI moat test on the same night
By Business & Technology Global · Mon Sep 07 2026
Oracle headquarters campus — the cloud-infrastructure vendor facing an AI capex test this week. Wikimedia Commons.
Oracle and Adobe report earnings on the same evening Thursday, 10 September — a clean side-by-side test of two AI business models: one that must build physical capacity, and one that must prove generative tools deepen rather than hollow out its software moat.
Oracle’s problem: growth that eats cash
Adobe campus exterior — the creative-software company reporting AI subscription progress Thursday. Wikimedia Commons.
Demand is not Oracle’s obvious bottleneck. Customers have prepaid for, or directly supplied, around $75 billion of GPUs tied to large AI contracts, reducing some funding pressure. The investor question is whether backlog converts into revenue and free cash flow fast enough to justify today’s buildout.
Adobe’s problem: a moat AI can redraw
Rows of datacenter server racks — the physical plant behind AI cloud infrastructure spend. Wikimedia Commons.
Adobe starts from the opposite balance sheet: abundant cash, light physical spend. Generative tools attack the scarcity that once made specialist creative software essential. Adobe’s response has been to host outside models — including OpenAI’s — inside its workflows, betting the moat is the professional create-edit-approve-publish surface, not ownership of a single model.
Saxo strategist Ruben Dalfovo’s 7 September note put the contrast bluntly: Oracle needs AI demand to become big enough; Adobe needs AI creation not to become too easy.
