Oracle is bringing Google’s Gemini models into its enterprise software stack, the company announced this week, extending the multi-cloud strategy it has pursued with Microsoft into a second major partnership. The move gives Oracle customers access to Gemini through the same interfaces they already use for their databases and business applications.
The integration runs at two levels. Google’s models will be available on Oracle Cloud Infrastructure, letting companies run Gemini workloads on Oracle’s machines, and Oracle will weave the models into its application software, so that its enterprise customers can call on AI features without switching platforms.
The announcement completes the second leg of a strategy Oracle began when it struck its partnership with Microsoft. Under that agreement, Oracle’s database software runs on Microsoft’s Azure cloud, and the two companies sell each other’s services. The Google deal extends the same logic: Oracle wants to be the neutral platform on which every major AI model runs.
The partnership is significant for Google. Oracle’s customer base is concentrated in banking, retail, healthcare and other industries that are cautious about adopting new technology, and the deal gives Gemini a distribution channel it has lacked. Google has been selling enterprise AI through its own cloud, but its share of the market trails Microsoft and Amazon by a wide margin.
For Oracle, the deal is a hedge. The company has said it does not plan to build its own foundation models, a decision that leaves it dependent on others for the intelligence layer of its software. By partnering with both Microsoft and Google, Oracle ensures its customers can use whichever model they prefer without leaving the platform.
The company’s position in the AI market is unusual. Oracle is a late entrant to cloud computing, and its infrastructure business is far smaller than the leaders’. But its database dominance gives it a grip on the data that enterprise AI depends on, and its applications business reaches millions of corporate users.
The strategy has won a following on Wall Street. Oracle shares have climbed steadily as investors bet that the company can capture a share of AI spending without building the enormous infrastructure that competitors are funding. The Google deal reinforces that narrative: Oracle collects fees on usage without paying for the research that made the models possible.
The integration work is the hard part. Enterprise customers run their business processes on systems that were not designed for AI, and connecting models to data securely, without letting sensitive information leak into training sets, requires engineering that most vendors are still developing. Oracle’s pitch is that its deep integration with corporate databases gives it an advantage.
The competitive picture is getting more crowded. Microsoft sells OpenAI’s models to its enterprise customers, Amazon offers a catalog of models on its cloud, and Google now reaches Oracle’s base. The models themselves are becoming interchangeable, and the platforms that deliver them are where the profit will be made.
Oracle’s history explains the strategy. The company dominated database software for decades, then entered cloud computing late, under the direction of its founder, and spent heavily to catch up. Its cloud infrastructure business, while smaller than the leaders’, has grown quickly, and its database lock-in gives it an entry point that pure cloud providers lack.
The database moat is the reason the deals work. Oracle’s software runs the core systems of most large banks, retailers and government agencies, and those customers are now deciding how to use AI with the data those systems hold. By offering access to multiple models, Oracle positions itself as the neutral layer between corporate data and the AI providers.
The partnership’s second leg matters for Google as much as Oracle. Google has struggled to convert its research leadership into enterprise cloud sales, and its market share trails Microsoft and Amazon. The Oracle deal gives Gemini a channel into industries, from banking to health care, where Google has little direct presence.
The competitive response is already visible. Microsoft has deepened its own AI partnerships, and Amazon has built a model catalog on its cloud. The platforms are converging on the same strategy, offering every major model through every major cloud, and the differentiation is shifting to price, performance and the quality of the integration with enterprise data.
The customer pitch is straightforward. A bank running its core systems on Oracle can use Gemini for document processing, Microsoft’s models for customer service and its own models for internal reporting, all through the same contracts and the same security controls. Oracle’s sales force, which has relationships with nearly every large enterprise, is the distribution machine the partnership puts to work.
For Oracle’s customers, the practical effect is choice. A bank that runs its core systems on Oracle can now use Gemini, or Microsoft’s tools, or both, through the same contracts and the same support desk. Whether that makes Oracle the AI platform of record, or just another way to access someone else’s model, will be decided by the usage numbers in the coming quarters.
