Oracle vs Adobe: AI is rewarding scarcity and punishing uncertainty

Key takeaways

• Oracle is turning artificial intelligence demand into real cloud revenue, but the bill for building capacity remains enormous.

• Adobe’s artificial intelligence products are growing quickly, yet investors still want proof that adoption protects pricing power.

• The bigger lesson is simple: artificial intelligence creates value only when it improves economics or strengthens competitive advantages.

Oracle and Adobe reported earnings on 10 September 2026, but they offered investors two very different artificial intelligence stories.

Oracle, once mainly known for databases, is spending heavily to become a major supplier of computing power. Adobe, the company behind Photoshop and Acrobat, already owns highly profitable software franchises but must defend them as AI makes content creation easier.

Oracle needs AI to become big enough to justify the investment. Adobe needs AI to avoid making its products too easy to replace. The earnings did not settle either investment case. They did, however, make the questions much clearer.

Oracle finally shows the receipt

Oracle’s strongest answer came from its cloud infrastructure business, where revenue jumped 121% to 7.4 billion USD.

That matters more than the headline earnings beat. Oracle has spent the past year asking investors to accept huge spending today in exchange for much larger AI revenue tomorrow. This quarter offers evidence that tomorrow is starting to arrive.

Oracle added 850 megawatts of data-centre capacity during the quarter and signed more than 30 billion USD of additional AI cloud contracts. Remaining performance obligations, essentially contracted revenue still waiting to be recognised, reached 664 billion USD.

Demand therefore looks less theoretical.

But backlog is not cash. Oracle spent 28.5 billion USD on capital expenditure during the quarter, mostly on data-centre equipment. Customers increasingly provide prepayments or hardware themselves, which helps, but Oracle still needs to turn enormous infrastructure commitments into attractive returns.

The investor question has shifted from “is there demand?” to “can Oracle deliver it profitably?” That is progress. It is not the finish line.

Adobe has the users. Now it needs the economics

Adobe faces almost the opposite problem.

Revenue rose 13% to 6.76 billion USD, while annual recurring revenue from its newer AI-first products grew more than 150%. Adobe also passed one billion monthly active users across its creativity and productivity products.

Those are hardly signs of a business being abandoned.

Yet its fourth-quarter revenue outlook came in slightly below expectations. That small miss matters because investors are not really debating whether Adobe can add AI features. They are debating whether those features strengthen Adobe’s competitive advantage.

There is an important difference.

If AI makes Photoshop, Acrobat and Adobe’s other tools more useful, customers may spend more and become harder to lose. But if AI simply allows users to produce good-enough images, videos and documents through cheaper alternatives, Adobe may gain AI usage without gaining much pricing power.

That makes new chief executive Anil Chakravarthy’s task unusually clear when he takes over on 1 December: turn AI adoption into stronger economics, not simply more AI.

One technology, two very different tests

Oracle and Adobe illustrate why “AI exposure” tells investors surprisingly little.

Oracle sells something increasingly scarce: computing capacity. Its challenge is capital intensity. Adobe sells software with historically attractive margins. Its challenge is defending scarcity as AI lowers the cost of creating digital content.

For Oracle, investors should watch whether revenue and cash flow catch up with infrastructure spending. For Adobe, the better signal is whether AI products support stronger recurring revenue, customer retention and pricing.

The useful question is therefore not whether a company uses AI. It is whether AI makes the customer relationship stronger and the underlying economics better.

Risks worth watching

Oracle’s biggest risk is execution. Data centres are expensive, technically complex and dependent on chips, electricity and construction. Rising spending without equally strong revenue conversion would be an early warning sign.

Adobe faces competitive and leadership risk. AI-first rivals can move quickly, while Adobe is changing chief executive and still searching for a permanent finance chief. Strong AI usage combined with weak guidance or greater discounting would suggest adoption is not translating into enough economic value.

Investor playbook

• Follow conversion: for Oracle, compare contracted demand with actual revenue, cash flow and capital spending.
• Separate usage from monetisation: for Adobe, rapid AI adoption matters more if recurring revenue and pricing remain resilient.
• Watch the competitive advantage: ask whether AI makes each company harder to replace or simply more expensive to operate.

The two unfinished AI stories

Oracle and Adobe started this earnings season with opposite questions, and they finish it with those questions narrowed rather than answered.

Oracle has provided stronger evidence that AI demand is big enough. It still has to prove that building all this capacity produces attractive returns. Adobe has shown that customers are using its AI products. It still has to prove those products make customers willing to keep paying Adobe.

That is the more useful lesson for investors. AI is not one investment theme. It is a stress test for every business model. The eventual winners will not simply use more AI. They will turn it into stronger customer relationships, better economics and, eventually, more cash.