In its fiscal second quarter 2025 results, Oracle’s IaaS revenue jumped considerably, demonstrating faster cloud growth than major cloud service providers Amazon, Microsoft and Google.
Oracle CEO Safra Catz confirmed this source of opportunity in its earnings calls: “Record-level AI demand drove Oracle Cloud Infrastructure revenue up 52% in Q2, a much higher growth rate than any of our hyperscale cloud infrastructure competitors.” Oracle noted that its AI success is largely from training generative AI (GenAI) models, and enterprise use of AI cloud training is still only in its infancy. Oracle’s future in the cloud market might hinge on further enterprise interest in Oracle to train AI models with their own data.
Let’s look at some of the factors contributing to Oracle’s success in the cloud AI market and ask some vital questions about the company’s future.
Oracle’s AI advantages
The following are some of the key advantages Oracle has in the cloud market.
Enterprise expertise
Enterprises have put Oracle in the top four in terms of strategic planning influence for two decades, and it has often been ranked second only to IBM. This gives Oracle input into major enterprise IT decisions, and surely, further enterprise adoption of AI could be one such decision.
Authority in databases
Oracle’s database products are widely used in both the cloud and data centers. Training AI means providing models access to historical data. Oracle understands enterprise data security and sovereignty needs, which means it gains greater enterprise trust in those critical areas. The breadth of its database use provides a large base into which it can promote agent AI offerings, large enough to sustain growth for several years.
Better pricing
Enterprises say that Oracle’s cloud is less expensive than the major cloud providers and that their pricing favors AI training by lowering access and transfer charges between the cloud and data center. They admit that Oracle doesn’t have the same global hosting density as the top three cloud providers, nor does it have the same web service feature options, but the AI training mission doesn’t require these things. With Oracle, enterprises can build a self-hosted cloud model and train it by transferring data to the Oracle cloud and then moving the model back. Alternatively, they can run an open source model in the cloud.
Assessing Oracle’s future in the AI cloud market
Despite Oracle’s recent success, there is never a guarantee that it will last. To determine Oracle’s foothold in the AI cloud market, keep an eye on three areas: Oracle’s ability to stand out from competitors, enterprise commitment issues and Oracle’s room for innovation.
Can Oracle AI differentiate itself?
The first issue is the durability of Oracle’s differentiators, and this issue has two dimensions:
- Competition. Can other cloud provider services offer similar or even better pricing and performance in AI training? With larger resource pools, why should Oracle retain price and performance advantages?
- Credibility. Organizations like IBM have the following:
- A larger base of enterprises as major accounts.
- Similarly powerful database and application positioning.
- Positive reputations for AI initiatives.
Can Oracle fend off the competition by building a reputation as the credible name in GenAI development?
Will enterprises commit to proprietary AI models?
Another significant factor to consider is the dependence of Oracle’s cloud infrastructure success on AI training, which is related to which enterprises train their own models. Growth in public GenAI services from Oracle competitors entering the AI space will likely disrupt Oracle’s AI training niche. Today’s enterprises are not committed to building AI models since they are a moving target. New models and new features roll out every day. Will enterprises commit to proprietary AI models during a time of such rapid evolution or wait for market stability?
Will Oracle continue to innovate?
Even if enterprises embrace a significant commitment to training AI models and Oracle optimally takes advantage of the AI training opportunity, could that alone keep Oracle’s cloud infrastructure business alive? That would almost certainly require Oracle to push AI beyond training and make itself a leader in AI applications.
Oracle has a SaaS application business, but it grew at a modest 9% in its recent report. There is no indication that this business draws much on AI. In addition, unlike rival IBM, Oracle is not known as a broad and capable AI evangelist for enterprises. Furthermore, Oracle doesn’t mention any potentially explosive AI opportunities in its financial reports, only AI training. Without decisive action, Oracle’s ability to exploit AI is limited to a role in training, which is insufficient to sustain the kind of growth the company expects and make it a full alternative cloud provider for enterprises.
Editor’s note: The enterprise views presented are taken from comments enterprises have made to the contributor. Where the views refer to a period prior to 2023, they were offered to CIMI Corp. After that, they were offered to Andover Intel.
Tom Nolle is founder and principal analyst at Andover Intel, a consulting and analysis firm that looks at evolving technologies and applications first from the perspective of the buyer and the buyer’s needs. By background, Nolle is a programmer, software architect, and manager of software and network products. He has provided consulting services and technology analysis for decades.
