Oracle has set off a quiet but seismic shift in the cloud industry. At Oracle AI World 2025, the company confirmed to AIM that it is officially the first in the world to launch Multicloud Universal Credits, a single currency that allows enterprises to run AI and database workloads across AWS, Azure, Google Cloud, and Oracle Cloud Infrastructure (OCI) with complete commercial freedom.
While hyperscalers continued tightening their perimeters through proprietary interconnects and marketplace constraints, Oracle took the opposite route of breaking the cloud walls entirely.
And the person explaining this shift with rare clarity is someone who has seen Oracle’s evolution up close for nearly three decades, Maria Colgan, a 1996 Oracle veteran, widely regarded as one of the most influential database minds inside the company.
Today, as VP of product management for AI and mission-critical data, she leads the product direction for Oracle’s AI engines, Advanced Compression, Database In-Memory, and more. She calls herself a “database nerd,” and her SQLMaria blog is followed by engineers across the world.
Speaking exclusively to AIM, she laid out why this moment matters.
Freedom, Finally
Asked how Universal Multicloud Credits change how enterprises run AI and data across multiple clouds, Colgan said the model gives customers true freedom to run workloads anywhere their data resides, without worrying about costs, movement, or cloud-specific limitations.
“It definitely does change the game. It’s going to open up the freedom that our customers are used to with the Oracle Database, to allow them to run it anywhere they want and wherever their data may be.”
“That Universal Credit… opens up that playground for them to be able to leverage the data wherever it exists, without having to worry about ingress fees and moving data around to get that holistic view of their business,” she added.
This is where Oracle’s strategy diverges sharply from AWS, Azure, and GCP. When asked if Oracle was turning its biggest competitors into distribution channels, Colgan reframed the question with Oracle’s long-time philosophy. She said Oracle sees them as partners working together to bring the best database and AI technologies to customers, enabling agentic workloads to run anywhere without concerns about cost or data movement.
Further, she confirmed, saying that Oracle is currently leading the market with this model.
“One of the first initiatives… to give customers that level of freedom. Mr Ellison and Oracle have always had that vision of freedom for our customers. I do think we are the first in the industry to do that, and it’ll change the way people are going to manage their expenses in cloud going forward,” she added, saying, “That’s why people are excited about this, to access everything, right?”
Clearly, Oracle is really good at naming its products
What Oracle has done commercially is as important as what it has done technically. Karan Batta, senior vice president at Oracle Cloud Infrastructure, broke down the system with characteristic humour. “Clearly, Oracle is really good at naming their products.”
“It’s a single currency that allows you to purchase any Oracle Cloud or Oracle product,” he added, saying, “You can use that same currency to consume database services on any cloud.”
This is the first time enterprises are being offered commercial portability, not just technical portability, across the four dominant clouds.
26ai: AI Where the Data Lives, and Not the Other Way Around
But Oracle didn’t stop with a commercial breakthrough. It also introduced its most ambitious database update in decades, Oracle Database 26ai.
Colgan explained why this matters.
“We’ve put AI into the data platform… not a bolt-on or add-on. Before, you had to wait for data to leave transactional systems, move downstream, and then use third-party tools for insights. By putting AI into the transaction processing systems, we can do real-time AI right there as part of those transactions,” she said
Where competitors run AI only inside analytics engines or lakehouses, Oracle is running it everywhere, including on mission-critical systems that run banks, telcos, and retailers.
This architecture becomes even more important in the age of agentic AI, where workflows require multi-step reasoning.
“With agentic AI, you’re allowing the AI to do reasoning, decision-making based on inputs and the environment. Because people have data across different clouds, being able to access and use that data in these multi-step workflows is going to be huge,” Colgan added.
She also highlighted the ongoing shift from public to private data as the true competitive moat. “The real moat is where you get the private data… being able to do that securely inside the Oracle Database is incredibly mind-blowing for customers,” she said.
Colagan said that everything inside 26ai, including vectors, embeddings, RAG, graph, spatial, and relational, is engineered to work together. “We’re not asking you to integrate those… we’re taking care of that on the backend, simplifying things and giving you that unified security approach, high availability, and scalability,” she concluded.
The post Oracle Just Solved Multicloud’s Biggest Problem appeared first on Analytics India Magazine.
