As of about 6:00 a.m. ET on Friday, December 12, 2025, U.S. big tech stocks are mixed in premarket trading—with investors juggling two competing narratives: AI demand remains powerful, but the cost of building it is rising fast, and markets are getting pickier about who can turn that spending into profits.
The early mood is being set by Oracle’s continued slide after a capex-heavy outlook, a Broadcom forecast that’s upbeat on revenue but cautious on margins, and a fresh round of policy and regulatory headlines spanning Nvidia’s China chip sales, Google’s EU antitrust exposure, and App Store rules in the U.S. [1]
Below is what’s moving the “Magnificent Seven” and other AI-linked mega-cap tech names before the opening bell—and what analysts are watching as the market heads into Friday’s session.
Premarket snapshot: Magnificent Seven and key AI infrastructure names
Premarket moves can change quickly, but around 6:00 a.m. ET the tape shows a clear split: AI infrastructure names tied to “capex shock” are under pressure, while some mega-caps are holding steadier.
Selected premarket indications (price / % vs. prior close):
- Oracle (ORCL): ~$198.85, -10.85%
- Alphabet (GOOGL): ~$312.43, -2.42%
- Nvidia (NVDA): ~$180.93, -1.57%
- Broadcom (AVGO): ~$406.37, -1.50%
- Tesla (TSLA): ~$446.89, -1.01%
- Amazon (AMZN): ~$230.28, -0.65%
- Apple (AAPL): ~$278.03, -0.24%
- Meta (META): ~$652.71, +0.40%
- Microsoft (MSFT): ~$483.47, +1.06%
- AMD (AMD): ~$221.43, ~flat
The biggest story this morning: Oracle’s “AI capex shock” rattles the trade again
Oracle is the name big tech traders are watching first because it’s become a proxy for the broader AI buildout—and because it sits at the center of the industry’s most watched AI infrastructure partnership.
On Thursday, Oracle shares sank about 13%, and Reuters reported the selloff was driven by a combination of massive spending and weaker-than-expected forecasts, stoking fresh doubts about how quickly AI investments will pay back. [2]
Why Oracle’s update hit so hard
According to Reuters, Oracle:
- Flagged a higher capex path, saying its fiscal 2026 spending is now expected to be $15 billion higher than its September estimates. [3]
- Missed a closely watched metric for future cloud contracts and issued a third-quarter revenue growth forecast below Wall Street estimates. [4]
- Has burned around $10 billion in cash in the first half of its fiscal year amid AI investments. [5]
- Has around $100 billion in debt, and investors have been increasingly focused on the credit implications of the AI buildout. [6]
- Saw at least 13 brokerages slash price targets after the results, even as some analysts argued the capex cycle is necessary to meet demand. [7]
Oracle’s pullback isn’t just a single-stock story. Reuters noted other AI-linked chip and infrastructure names sold off alongside it, underscoring how tightly correlated the AI complex remains when the market starts questioning “returns on investment” timelines. [8]
A broader “AI bubble” debate is back in focus
This is the recurring theme: the market is still buying the AI demand story, but it’s increasingly skeptical about who captures the profit and when.
A Reuters analysis this week highlighted how AI infrastructure is increasingly being financed with debt, and cited a UBS estimate that AI data-center and project-financing deals surged to $125 billion so far this year, up from $15 billion in the same period of 2024. [9]
That shift matters because it changes the financial math around AI expansion—especially if growth slows, power costs rise, or deployment timelines slip.
Broadcom’s forecast: AI demand is strong—profitability is the near-term trade-off
If Oracle is the “spending risk” story, Broadcom is the “demand is real, but margins are under pressure” story.
Reuters reported Broadcom projected first-quarter revenue above Wall Street estimates, but warned that gross margins would dip due to a higher mix of AI revenue—sending shares down in extended trading. [10]
Key takeaways from Broadcom’s update:
- Broadcom forecast quarterly revenue of about $19.1 billion vs. an analyst average estimate of $18.27 billion (LSEG data via Reuters). [11]
- The company expects gross margin down ~100 basis points sequentially, tied largely to the AI revenue mix. [12]
- CEO Hock Tan referenced a $73 billion backlog expected to ship over the next 18 months. [13]
- Broadcom said its AI semiconductor revenue (custom chips + networking for AI data centers) is expected to double to $8.2 billion in fiscal Q1. [14]
- Analysts also flagged customer concentration risk (a handful of large buyers) and warned that systems sales can carry lower margins as mix shifts. [15]
In other words: AI demand isn’t the problem—the market is now focused on profit quality, margin trajectory, and whether AI product mixes dilute profitability before the next wave of monetization arrives.
Macro backdrop: Nasdaq futures cautious after Oracle; Fed-cut week still shaping sentiment
Overnight, Reuters reported Nasdaq futures were down about 0.1% and S&P 500 e-mini futures were little changed, with Oracle and Broadcom weighing on tech sentiment. [16]
The macro context matters for mega-cap tech because these stocks tend to react strongly to moves in rates, the dollar, and the growth outlook:
- Reuters noted the market had to “move fast” this week after the Federal Reserve cut interest rates but delivered a less dovish outlook than some investors expected. [17]
- The U.S. Dollar Index was cited near a two-month low (98.37 in that report), which can be supportive for multinationals and risk assets at the margin. [18]
- The 10-year Treasury yield was cited around 4.1586% in the same overnight wrap. [19]
This combination—rate-cut tailwind with AI ROI doubts—is a big reason premarket tech trading feels “twitchy”: investors want growth, but they’re increasingly demanding proof that AI spending converts into earnings power.
Regulation and geopolitics: a busy headline tape for Big Tech
Alongside earnings-driven volatility, Friday’s premarket is also digesting a dense set of regulatory signals—especially around chips, platforms, and AI governance.
Nvidia: China chip sales face renewed scrutiny in Washington
A Reuters report said Democratic Senator Elizabeth Warren called for Nvidia CEO Jensen Huang and the Commerce Secretary to testify after President Donald Trump announced plans to greenlight sales of Nvidia’s H200 AI chip to China. [20]
Nvidia emphasized that sales would still require a U.S. government license, and the report noted the debate is intensifying amid national-security concerns and allegations of chip smuggling. [21]
For markets, this is the recurring Nvidia variable: export rules can reshape demand and pricing quickly, even when overall AI appetite remains strong.
Google/Alphabet: EU fine risk under the Digital Markets Act
Reuters reported Google is expected to face an EU antitrust fine next year for not doing enough to comply with EU rules that prohibit favoring its own services in search results, according to people familiar with the matter. [22]
The report also highlighted:
- The case involves alleged self-preferencing of services like Google Shopping, Hotels, and Flights. [23]
- DMA violations can lead to fines up to 10% of global annual turnover. [24]
- The issue is separate from another investigation involving Google Play, Reuters said. [25]
This helps explain why Alphabet is one of the weaker mega-caps premarket even as the broader group is mixed. [26]
Apple: App Store competition ruling + child online safety fight
Apple has two major policy arcs on investors’ radar:
1) Epic Games / App Store antitrust:
Reuters reported Apple won a partial reversal of contempt-related sanctions in the Epic Games case, but the appeals court upheld much of the contempt finding and an earlier injunction. The ruling gives Apple an opening to argue for a “reasonable commission” on certain off-platform purchases, while still keeping pressure on App Store practices. [27]
2) Age verification and privacy:
Separately, Reuters reported Apple CEO Tim Cook met U.S. House members to push back on federal legislation that could require app stores to authenticate users’ ages—arguing it could lead to broad collection of sensitive personal data. [28]
For Apple (and Google), these debates aren’t just “politics”—they can influence the long-term economics of the app ecosystem and services revenue durability.
Microsoft: cloud licensing lawsuit in the UK
Reuters reported Microsoft is fighting a 2.1 billion-pound (~$2.8 billion) UK lawsuit alleging it overcharged thousands of British businesses to use Windows Server software on rival cloud platforms such as AWS and Google Cloud. [29]
Regulators in the UK, EU, and U.S. have been examining cloud competition issues, Reuters noted, putting cloud economics and licensing practices back into focus. [30]
Amazon: retail logistics push with one-hour pickup concept
On the commerce side, Reuters cited a Business Insider report that Amazon is developing a “rush” pickup service that could let customers collect orders from stores within an hour, with a pilot potentially as soon as early 2026. [31]
While not as market-moving as Oracle/Broadcom, it reinforces Amazon’s ongoing push to compete on speed and convenience—a key part of the investment case heading into 2026.
AI governance: federal procurement and state regulation pressure
Two Washington-linked AI headlines are also in the mix:
- Reuters reported the U.S. government is set to require AI vendors to measure political bias as part of selling tools to federal agencies. [32]
- Reuters also reported President Trump threatened federal funding for states over AI regulations. [33]
Investors typically treat these as longer-cycle issues, but they can affect product roadmaps, compliance costs, and public-sector AI adoption narratives—especially for Microsoft and other enterprise AI platforms. [34]
The bigger “AI buildout” constraint: power and financing
A quieter but increasingly market-relevant driver of big tech’s AI story is infrastructure constraints—especially power.
Reuters Events reported big tech has shifted to an “all of the above” strategy to power AI, expanding beyond renewables into gas-fired and nuclear arrangements as data-center demand accelerates. [35]
Notable data points from that Reuters report include:
- U.S. utilities’ power supplies to data centers are forecast to jump 22% this year to 61.8 GW, reaching 134.4 GW by 2030, citing an S&P Global outlook. [36]
- The International Energy Agency data cited in the piece suggests gas-fired power remains a major source for U.S. data centers due to the need for 24/7 firm power, with clean power expected to play a larger role later as capacity expands. [37]
This matters to stock investors because “AI capex” isn’t just chips and servers—it’s also power procurement, grid access, and long-duration infrastructure financing, which can pressure margins and reshape timelines.
What to watch when the bell approaches
Heading into the open, here are the pressure points most likely to drive big tech direction during Friday’s session:
- Does Oracle stabilize—or keep dragging the AI complex?
Oracle’s capex path and cloud outlook are acting like a market-wide “AI confidence test.” [38] - Broadcom’s margin message vs. revenue momentum
Investors will debate whether the margin dip is a temporary “mix shift” or a signal that AI hardware economics are getting more competitive. [39] - Nvidia’s policy risk premium
Any further developments on China export licensing and Washington scrutiny can move Nvidia quickly—often spilling into the broader semiconductor group. [40] - Alphabet’s DMA exposure
Regulatory headline flow out of Europe remains a real variable for Alphabet’s multiple—especially around search product design and monetization. [41] - The “capex + power + debt” theme
AI’s next leg increasingly depends on infrastructure financing and reliable power—topics now regularly appearing in market analysis, not just energy trade journals. [42]
Bottom line
Big tech remains the market’s center of gravity—but Friday’s premarket shows investors are shifting from simply buying “AI winners” to demanding clearer answers on ROI, margins, capex discipline, and regulatory risk. Oracle’s stumble and Broadcom’s margin warning are a reminder: AI demand can be booming while the market still punishes uncertainty about the payoff. [43]
This article is for informational purposes only and is not financial advice. Premarket prices are indicative and can change rapidly before the opening bell.
References
1. www.reuters.com, 2. www.reuters.com, 3. www.reuters.com, 4. www.reuters.com, 5. www.reuters.com, 6. www.reuters.com, 7. www.reuters.com, 8. www.reuters.com, 9. www.reuters.com, 10. www.reuters.com, 11. www.reuters.com, 12. www.reuters.com, 13. www.reuters.com, 14. www.reuters.com, 15. www.reuters.com, 16. www.reuters.com, 17. www.reuters.com, 18. www.reuters.com, 19. www.reuters.com, 20. www.reuters.com, 21. www.reuters.com, 22. www.reuters.com, 23. www.reuters.com, 24. www.reuters.com, 25. www.reuters.com, 26. www.reuters.com, 27. www.reuters.com, 28. www.reuters.com, 29. www.reuters.com, 30. www.reuters.com, 31. www.reuters.com, 32. www.reuters.com, 33. www.reuters.com, 34. www.reuters.com, 35. www.reuters.com, 36. www.reuters.com, 37. www.reuters.com, 38. www.reuters.com, 39. www.reuters.com, 40. www.reuters.com, 41. www.reuters.com, 42. www.reuters.com, 43. www.reuters.com
