NEW YORK, April 23, 2026, 09:45 EDT
- Oracle expanded its Google Cloud AI database product lineup on Wednesday. Still, the stock slid Thursday, caught in a broader software sector downturn.
- Morgan Stanley trimmed its price target on Oracle to $207, citing questions about GPU-as-a-service, funding, and profitability.
- Whether Oracle’s rapid cloud expansion can offset its hefty $50 billion capital spending target for fiscal 2026 remains the stock’s key question.
Oracle stock slipped Thursday, coming off a fresh extension of its Google Cloud tie-up, as a broad software slump overshadowed the newest AI announcement. Premarket moves pointed to a loss as steep as 3.8%, with shares ticking down to $180.40 from a $187.50 close the day before.
This shift is significant: Oracle’s valuation now leans heavily on AI infrastructure alongside its legacy software sales. Demand is surging. Still, investors are zeroing in on margins, debt load, and the expense tied to expanding GPU-as-a-service—Oracle’s cloud-based rentals for the graphics chips behind AI model training and inference.
Oracle on Wednesday announced several updates for its Oracle AI Database@Google Cloud, including a new AI agent built for Gemini Enterprise, expanded regional access, and enhanced integration with Google’s data tools. Availability now stretches to 15 regions, according to the company.
With the new agent, customers can pull up Oracle data just by asking questions in plain English—no SQL needed. “Access, understand, and act” on enterprise data: that’s the goal, according to Nathan Thomas, senior vice president at Oracle Cloud Infrastructure. He emphasized that controls remain at the database layer. Oracle Blogs
At its Cloud Next conference in Las Vegas, Google framed the partnership as a step in its wider push into the enterprise market. “Customers need a simple, trusted way to work with business data through intelligent agents,” said Satish Thomas, Google Cloud’s vice president for applied AI and platform ecosystem. Reuters noted that Google is seeking to establish a software moat around Gemini Enterprise, aiming to capture a bigger slice of corporate AI budgets as it faces off against Microsoft and Amazon. Oracle
Oracle’s multicloud push isn’t stopping with Google. Just last week, the company announced plans to boost private, high-speed connections between Oracle Cloud Infrastructure and AWS, deepening its existing network links to AWS, Google Cloud, and Microsoft Azure. The move is designed to give customers more flexibility to run workloads across several cloud providers.
Wall Street’s opinions remain split. On Thursday, Morgan Stanley’s Keith Weiss lowered his price target on Oracle to $207 from $213, maintaining an Equal Weight rating. Weiss flagged lingering questions about GPU-as-a-service, financing, and the margins attached to that rapid-growth segment.
Caution hasn’t let up on the stock for months. After Oracle’s March-quarter numbers, Hargreaves Lansdown’s Matt Britzman flagged the financing debate as something that’s “not going away anytime soon.” Customer prepayments and bring-your-own-hardware deals could boost Oracle’s future revenue commitments without sticking the company with the entire bill, he added. Back then, Reuters pointed out remaining performance obligations—contracted revenue still unbooked—soared 325% to $553 billion. Reuters
Oracle continues to post robust results. Cloud revenue climbed 44% to $8.9 billion in the March quarter, while cloud infrastructure soared 84% to $4.9 billion. The company is projecting total cloud revenue growth of 46% to 50% for the current quarter and is sticking with its forecast of $50 billion in capital expenditures for fiscal 2026.
The scale of investment around AI infrastructure just keeps widening. Earlier this month, Reuters said Related Digital was closing in on $16 billion in financing for an Oracle-run data center project in Michigan—part of Stargate, the stateside AI buildout with OpenAI and others in the mix. Last week, Bloom Energy announced it will deliver as much as 2.8 gigawatts of fuel-cell capacity to Oracle, aimed at keeping up with surging AI power demand.
The risk stands out. IBM’s sluggish software numbers and ServiceNow flagging deal delays in the Middle East sparked Thursday’s sharp software retreat, a fresh reminder of how quickly investors can sour on AI bets if earnings don’t follow the hype—or if expenses climb instead. UBS strategists now argue the hurdle isn’t just spinning an AI narrative but showing it actually delivers for “products, workflows, and returns.” Investing.com
Oracle finds itself in a tight spot. While it’s expanding partnerships with Google Cloud and AWS, the company’s shares are still moving on questions about whether AI demand can keep pace with the soaring costs of scaling up new capacity. Investors lately have shown a clear preference for chip stocks rather than software.
