Oracle integrates Google AI into enterprise apps, shares jump 8%
Oracle is embedding Google's Gemini AI into its core business software, a move that rewards investors and signals a broader corporate shift away from single-vendor AI lock-in.
Oracle shares climbed 8.4% to $127.64 on Thursday after the company announced it would integrate Google’s Gemini AI models into its enterprise applications. The technology will power embedded AI across Oracle Fusion and NetSuite, joining existing models from Cohere and Meta.
Customers will choose between two Google models depending on the task. Gemini 3.1 Flash-Lite is built for cheap, high-volume operations, while Gemini 3.5 Flash handles complex reasoning, video, and presentations. Both will run inside Oracle AI Agent Studio for Fusion Applications.
The deal highlights a decisive shift in enterprise technology. Rather than forcing clients to use a proprietary AI, Oracle is adopting a multi-model approach. “Organizations need the flexibility to choose the AI model best suited to each problem,” said Chris Leone, Oracle’s applications-development chief.
The partnership is notable because Oracle and Google are fierce rivals in cloud computing. It also bypasses OpenAI, Oracle’s roughly $300bn infrastructure partner, at the application layer. For Google, the arrangement provides instant access to NetSuite’s more than 44,000 corporate customers. For Oracle, it offers a new selling point as it spends heavily on AI data centres while managing a substantial debt load.
The liability question
However, embedding autonomous agents into critical business workflows carries significant risks, particularly around accountability. If an AI agent executes a flawed transaction, determining liability remains unresolved. “It sounds good, but be cautious,” said Gartner analyst Balaji Abbabatulla. “It doesn’t necessarily look as glittery as it sounds. There are challenges under the hood.”
Oracle points to its monitoring and audit tools to mitigate these risks, but analysts remain skeptical that any vendor has fully solved the problem. For businesses running these systems, the deal offers operational flexibility, but the hard questions about governing autonomous AI in the workplace are still open.
