Oracle’s Second Act: Losing Cloud, Winning AI | Blog

For most of the past 15 years, Oracle looked like the giant that sat out the future. 

• In hyperscale cloud, it delayed while AWS and Microsoft Azure raced ahead, capturing developers and enterprises with global capacity and aggressive pricing 

• In Software-as-a-Service (SaaS) and enterprise applications, Salesforce defined the Customer Relationship Management (CRM) era, ServiceNow redefined workflow automation, and Workday became the Human Capital Management (HCM) reference point. Oracle’s own portfolio of Fusion and NetSuite, also acquired modules, but never achieved the same category dominance 

• In leadership narrative, Larry Ellison came to be remembered more as the billionaire who owned an island and hobnobbed with Elon Musk, than as the technologist shaping the next wave of enterprise IT 

Reach out to discuss this topic in depth.  

By the mid-2010s, Oracle was acknowledged as powerful but outdated, a company milking legacy databases while others seized the future. Then, in September 2025, everything changed. 

The chart below tells the story with brutal clarity: for more than a decade, Oracle’s stock barely moved, drifting sideways while rivals defined the future. Then, starting in 2020, the line began to slope upward, and in September 2025 it went vertical, a once-in-a-generation repricing that turned Oracle from an afterthought into one of the most valuable technology companies on the planet. 

Oracle Stock Price (2010-2025). Source: Macrotrends 

Oracle in September 2025: Five headline facts 

• $455 billion backlog, up 359 percent year over year — proof of massive demand already contracted 

• A $300 billion deal with OpenAI — a historic agreement positioning Oracle at the center of Artificial Intelligence (AI) infrastructure 

• Capital expenditure raised to $35 billion — a clear signal of intent to scale data centers, Graphics Processing Units (GPUs), and interconnects 

• AI cloud revenue roadmap to $144 billion by 2030 — a public trajectory forcing investors to treat hypergrowth as baseline 

The market believed, and Larry Ellison briefly became the world’s richest man, leapfrogging his friend Elon Musk. 

The lost S-curves: How Oracle fell behind 

Oracle’s comeback is striking because of how thoroughly it missed the last two cycles: 

1. Delay in hyperscale buildout. Oracle clung to on-premise roots while its rivals blanketed the globe with data centers. That hesitation meant developers never defaulted to Oracle Cloud 

2. Fragmented enterprise applications story. While Salesforce redefined CRM, ServiceNow dominated workflows, and Workday owned HCM, Oracle never secured comparable category leadership. Its applications portfolio was broad but never cohered into a unifying movement 

3. Positioning drift. For too long, Oracle read as powerful but irrelevant. Ellison’s image reinforced the sense that the company was extracting rents from the past rather than inventing the future 

The next S-curve: What Oracle got right despite the misses 

Even while Oracle was missing cloud and SaaS peaks, it was putting together the pieces,  both technical and strategic, that made its September 2025 surge believable: 

• Industry depth. The Cerner acquisition opened Oracle into healthcare’s most regulated workflows, setting groundwork for clinical, financial, and operational systems with data, AI, and compliance all embedded 

• Enterprise workflow knowledge. Fusion Applications gave Oracle a cloud-compatible, process-oriented footprint in Enterprise Resource Planning (ERP), HCM, Customer Experience (CX) and Software Configuration Management (SCM), not leading in mindshare always, but deeply rooted in how big enterprises run 

• Regulatory and geographic positioning. Government/regulation matter: Oracle locked contracts with federal and defense agencies, negotiated major GSA pricing deals, and offered up discounts to become the low barrier for federal IT modernization 

• Political signaling and leadership alignments. Ellison’s visible engagement with the Trump administration via Stargate; public White House appearances with AI luminaries; positioning as a partner to government AI / security modernization, not just selling, but helping shape policy and infrastructure 

Those are the moves that underwrote the pivot, not least because building infrastructure without political, regulatory, enterprise alignment nearly always fails at scale. But the pivot itself was critical 

The pivot: A high-stakes infrastructure strategy 

Starting around 2022, Oracle reoriented itself around a single, audacious thesis: AI will be the most compute-hungry wave in history, and the winners will be those who commit capacity ahead of everyone else. 

• Mega-contracts as proof of demand, something that analyst could project future accruals on 

• Capital as the signal. Raising FY26 capex to $35B was not a financial footnote; it was the declaration that Oracle was prepared to match AWS and Microsoft in steel, silicon, and megawatts 

• Integration with the enterprise base. Unlike the hyperscale natives, Oracle tied its AI cloud narrative back to its existing dominance in enterprise data 

• Public accountability. By publishing a roadmap to $144B in AI cloud revenue by 2030, Oracle forced analysts to treat its trajectory as a baseline scenario, not as wishful thinking 

• Move to a two-Chief Executive Officer (CEO) structure fits this same logic. By handing Clay the fast-growing infrastructure and AI business, and Mike Sicilia the slower but steadier applications and industries, Oracle made its leadership map mirror the very split driving its second act. 

Why the market believes the Oracle (pun intended) this time 

This time, investors aren’t buying a story. They’re reacting to numbers with real gravity: 

• Contracts you can’t dismiss. $455B in backlog is enforceable. Analysts can’t wave it away; they must incorporate it into models 

• An arms race that rewards capacity. AI demand is constrained by GPUs, power, and data center real estate. Oracle moved early to secure supply 

• Execution already visible. Customers are running production AI workloads on OCI today. This isn’t speculative, it’s live infrastructure 

The return of the king? 

Oracle missed cloud. It missed SaaS and workflow automation. For years it was shorthand for a company that once mattered but no longer led. 

But September 2025 has put a question mark on that perception. Oracle is now one of the few firms with the contracts, the capital, and the political positioning to power the AI boom. And yet, the real test still lies ahead. 

• Can it turn a $455B backlog into revenue at the velocity Wall Street now demands? 

• Can it defend ground once AWS, Microsoft, and Google unleash their counterpunches with custom chips and ecosystem lock-in? 

• And can it manage the non-technical bottlenecks, power, regulation, federal scrutiny, where politics matter as much as GPUs? 

• And can a two-CEO model, which almost never works, avoid becoming a distraction just when Oracle needs singular focus? 

This comeback wasn’t just about infrastructure. Oracle’s bets on healthcare, enterprise workflows, and sovereign cloud gave it a base. Its federal contracts, GSA deals, and Ellison’s public alignment with Trump’s AI agenda showed an instinct for political leverage that few Silicon Valley peers dared. Those choices, quiet for years, became accelerants when AI exploded. 

These are not abstract questions. They will determine whether Oracle’s surge is remembered as a genuine reinvention or a spectacular head-fake. 

If you enjoyed reading this blog, check out Cloud Control Has A Capital: Jurisdiction Is The New Fault Line In Cloud Strategy | Blog – Everest Group, which delves deeper into other topics regarding cloud. 

Reach out to Abhishek Singh ([email protected]), to discuss more in depth about our insights and offerings at Everest Group.