Updated: Thursday, December 18, 2025 — 10:07 a.m. ET
AI stocks are trying to regain their footing this morning after a volatile 48 hours that put the entire “AI infrastructure” narrative back under the microscope. The tone shift is being driven by two forces pulling in opposite directions:
- A confidence boost from Micron’s bullish outlook, reinforcing that AI demand for high-bandwidth memory (HBM) is still running hot.
- A reality check from ongoing questions about how easily the market can finance the next wave of mega–data center projects—questions that flared around Oracle and rippled across semiconductors and AI-linked infrastructure names. [1]
Below is what matters for investors watching AI chip stocks, AI cloud stocks, and the broader AI supply chain today (12/18/2025)—with the key forecasts and analyses shaping price action.
The macro backdrop: Inflation cooled, giving growth stocks breathing room
A softer-than-expected U.S. inflation read is helping relieve pressure on high-multiple tech and AI stocks—especially those sensitive to interest-rate expectations.
The Consumer Price Index (CPI) rose 2.7% year over year in November 2025, while core CPI (excluding food and energy) rose 2.6% over the same period, according to the U.S. Bureau of Labor Statistics. [2]
Markets interpreted the data as supportive of the idea that rate cuts could return to the conversation next year—an important tailwind for long-duration growth themes like AI. [3]
At the opening bell, U.S. indexes were higher, reflecting that improved macro mood. [4]
Micron (MU) steals the spotlight: AI memory demand is outpacing supply
If there’s one headline moving “AI stocks” today, it’s Micron.
Micron surged in early trading after delivering a significantly stronger-than-expected profit forecast, with management pointing directly to tight supply and strong pricing for HBM—one of the most critical components for training and running large AI models in data centers. [5]
A few details from today’s coverage that investors are latching onto:
- Micron is one of the three major global HBM suppliers, alongside Samsung and SK Hynix—making its commentary a bellwether for the entire AI compute stack. [6]
- The company signaled supply tightness could persist beyond 2026, and indicated it may only be able to meet roughly half to two-thirds of demand from key customers. [7]
- Separate coverage highlighted Micron’s upside surprise on results and guidance, with revenue and earnings exceeding expectations and bullish forward projections reinforcing the AI-memory boom narrative. [8]
Why this matters for AI stocks broadly: HBM is effectively a “throughput limiter” for AI servers. When memory supply is tight and pricing is strong, it can mean (1) high profits for memory makers, and (2) continued urgency—sometimes desperation—among hyperscalers and AI labs to secure compute capacity.
Nvidia (NVDA), AMD (AMD), Broadcom (AVGO): Micron calms “AI spend slowdown” fears—at least for now
Micron’s read-through helped lift the rest of the AI chip complex in premarket/early trading, with Nvidia and other names rebounding after a pullback fueled by infrastructure financing headlines.
One widely cited dynamic this morning: Micron’s results made it harder to argue that AI spending is abruptly rolling over—at least at the component level where real orders are showing up. [9]
Adding fuel to the bull case, an AI infrastructure player announced plans to deploy more than 2,300 Nvidia Blackwell GPUs at a California facility as part of a broader buildout plan—another data point suggesting the AI build cycle is still expanding. [10]
At the same time, investors should recognize what today’s bounce does not resolve: the market is still debating whether 2026–2028 infrastructure plans can be financed smoothly—especially when deals start stacking into the tens of billions.
Oracle (ORCL) and the “AI financing” problem: the story that rattled the trade
The biggest overhang on AI stocks this week has been financing risk—not demand risk.
Oracle’s planned Michigan data center project faced fresh uncertainty after a key financial backer, Blue Owl Capital, withdrew from the deal, according to reporting that emphasized concerns over lease terms and the broader stress investors are placing on debt-heavy AI infrastructure expansion. [11]
That headline mattered beyond Oracle because it hit the AI market at a sensitive point: investors are increasingly asking whether the next leg of AI growth is constrained not by chips, but by capital structure—who funds the data centers, at what returns, and under what lease economics.
Reuters commentary this morning captured the broader market psychology: worries about the “stratospheric” sums pouring into AI have revived bubble talk, with Oracle placed near the center of that anxiety. [12]
How traders are reading it today:
Micron’s strength supports the “demand is real” case, but Oracle’s situation keeps the market focused on the “who pays for it?” question—especially for multi-gigawatt projects.
OpenAI’s mega-fundraise chatter: bullish for suppliers, but raises the stakes
Another story shaping the AI-stock narrative today is the scale of private capital being discussed at the frontier-model level.
Reuters reported that OpenAI has been in preliminary discussions about raising tens of billions of dollars—potentially up to $100 billion—at around a $750 billion valuation, citing The Information. The report also noted how the industry’s appetite for compute is escalating, even as investors remain cautious about long-term returns on massive AI spending. [13]
For public AI stocks, the implications cut both ways:
- Bullish read-through: more OpenAI capital can translate into more spending on GPUs, memory, networking, power, and cloud capacity (good for parts of the AI supply chain).
- Risk read-through: the bigger the numbers get, the louder the questions become about ROI, margin durability, and whether the financing environment can keep up.
The “AI infrastructure” trade widens: Core Scientific (CORZ) and Hut 8 (HUT) jump on AI compute deals
Not all “AI stocks” today are traditional software or semis. Some of the most aggressive pivots are coming from crypto-era infrastructure players repurposing power and real estate for AI workloads.
Core Scientific (CORZ): upgrade on HPC/AI opportunity
Core Scientific moved after a high-profile analyst upgrade pointed to its power pipeline and potential to secure additional high-performance compute (HPC) lease agreements beyond existing arrangements. [14]
Hut 8 (HUT): a long-term AI data center leasing deal
Hut 8 surged after announcing a $7 billion, 15-year deal tied to developing high-capacity data center infrastructure, involving Fluidstack and Anthropic, with 245 megawatts of computing power in the initial build and additional expansion rights. [15]
These moves matter because they reinforce a key theme for 2026: AI is increasingly a “power and megawatts” story, not just a “chips and models” story.
Forecast watch: Data center capex keeps accelerating into 2026
One of the most important “today” forecasts for AI stocks isn’t a single-company price target—it’s the capex trajectory of the firms building the AI factories.
A Dell’Oro Group update said worldwide data center capital expenditures increased 59% year over year in 3Q 2025, marking the eighth consecutive quarter of double-digit growth, and noted that the “Top 4” U.S. cloud providers—Amazon, Google, Meta, and Microsoft—have continued to raise data center capex expectations for 2025. The update also said the capex growth outlook was raised through 2026 and referenced accelerated server spending tied to Nvidia’s Blackwell ramp and custom accelerator platforms. [16]
This is the core tension in AI stocks right now:
- Spending momentum remains strong and broadening (bullish for the supply chain). [17]
- Yet the market is increasingly sensitive to any sign that financing terms, lease economics, or balance-sheet stress could slow the buildout (bearish for high-multiple AI names if sentiment turns). [18]
What to watch next today
Here are the catalysts most likely to keep AI stocks moving after the initial morning reaction:
- Follow-through in semis beyond Micron: Does strength spread from memory into GPUs, networking, and equipment—or does it stay “MU-specific”? [19]
- Any additional clarity around Oracle’s data center financing path: The AI market is treating this as a live test case for the next wave of mega-project funding. [20]
- The “AI capital cycle” narrative: OpenAI’s fundraising chatter pushes expectations higher, which can help suppliers—but also raises the bar for proving returns. [21]
- IPO pipeline for AI-linked names: Nasdaq has pointed to improving conditions for large IPOs and highlighted AI-adjacent names that have recently come public, a reminder that the AI trade increasingly spans both public and private markets. [22]
References
1. www.reuters.com, 2. www.bls.gov, 3. za.investing.com, 4. www.reuters.com, 5. www.reuters.com, 6. www.reuters.com, 7. www.reuters.com, 8. www.barrons.com, 9. www.barrons.com, 10. www.barrons.com, 11. www.reuters.com, 12. www.reuters.com, 13. www.reuters.com, 14. www.investors.com, 15. www.barrons.com, 16. www.prnewswire.com, 17. www.prnewswire.com, 18. www.reuters.com, 19. www.barrons.com, 20. www.reuters.com, 21. www.reuters.com, 22. www.reuters.com
