AI Stocks on a Rollercoaster: Mega-Deals, Soaring Valuations & Bubble Jitters (Sept 25

Key Facts

Nvidia’s $100B AI Power Play: Chipmaker Nvidia announced plans to invest up to $100 billion in OpenAI (maker of ChatGPT), supplying it with advanced chips ^[1]. The blockbuster partnership underscores Nvidia’s dominance in AI – analysts estimate each gigawatt of AI data center capacity can mean ~$50 billion in revenue, making this deal potentially worth $500 billion in future sales ^[2]. Nvidia’s stock initially jumped on the news, buoying its valuation to record highs, though some gains faded amid broader tech profit-taking mid-week.
Oracle’s Cloud Bonanza – and Pullback: Enterprise giant Oracle stunned the market by snagging one of the largest cloud contracts ever: OpenAI is expected to buy $300 billion of cloud capacity from Oracle over ~5 years ^[3] ^[4]. Oracle is also in talks with Meta on a separate $20 billion AI cloud deal ^[5] ^[6]. These deals sent Oracle’s backlog of cloud orders soaring to $455 billion, prompting Oracle to raise $18 billion in debt to build data centers ^[7] ^[8]. Oracle’s stock hit multi-year highs on the AI euphoria, but has since slid ~5% on Sept. 25 amid investor profit-taking and concerns over its heavy spending and debt ^[9].
Market Jitters Hit AI High-Flyers: After a meteoric 2023–2024 run, AI-focused stocks showed signs of cooling. Major indexes Nasdaq and S&P 500 closed down ~0.5% on Sept. 25 – the third straight decline after hitting record highs earlier in the week ^[10]. Investors are growing wary of lofty valuations: even as hyperscalers like Microsoft, Alphabet (Google) and Amazon boost capital spending 62% this year on AI data centers ^[11], the Fed Chair warned tech equity prices appear “fairly highly valued” ^[12] ^[13]. Notably, Oracle fell 5.6% Thursday ^[14], Tesla (touted for its self-driving AI) sank over 4%, and Nvidia and Micron dropped around 1–3% amid a broader tech pullback ^[15] ^[16]. AMD held up better (even rising slightly) as it touts new AI chip traction, but it too faces skepticism over high inventory levels and a 41× P/E ratio ^[17].
AI Startups Strike It Rich: A wave of AI-focused startups is riding the investment boom. Cloud infrastructure firm CoreWeave – backed by Nvidia – reportedly IPO’ed this year and quickly inked $6.3 billion and $11.9 billion deals to provide AI cloud capacity to Nvidia and OpenAI, respectively ^[18] ^[19]. Scale AI, a data-labeling startup, got a $14.3 billion strategic investment from Meta for a 49% stake ^[20]. Even legacy chipmakers are drawing big backers: Intel secured a surprise $2 billion cash infusion from SoftBank as it struggles to catch up in AI chips ^[21], and Nvidia agreed to buy a 4% stake in Intel for $5 billion (contingent on Intel issuing new shares) ^[22]. Meanwhile, Amazon doubled its bet on OpenAI rival Anthropic with a fresh $4 billion investment to bolster its AI offerings ^[23]. These deals highlight how trillions of dollars are being poured into AI infrastructure across the industry.
Analysts Split: Boom or Bubble? Expert commentary is sharply divided on whether AI stocks are the opportunity of the decade or a brewing bubble. Bullish view: Top Wall Street analysts argue the AI revolution justifies the hype – Barclays noted a “wave of announcements” has unveiled over $2 trillion in planned AI spending, making even a $3–4 trillion industry by 2030 look “much more real” ^[24] ^[25]. Nvidia’s latest moves prompted price target hikes (Barclays now sees another +36% upside for NVDA stock ^[26]), and one UBS analyst projects Nvidia will dominate a $3–4 trillion AI market by 2030. Executives are similarly upbeat: Nvidia’s CEO calls AI “the new engine of the world economy,” and Oracle’s founder Larry Ellison boasts that surging AI cloud demand could make Oracle the world’s #2 software firm. Bearish view: Others warn the frenzy has outrun reality. A Reuters Breakingviews analysis by Edward Chancellor argues companies and investors are “trapped inside the bubble” of AI – pouring money into projects unlikely to earn adequate returns ^[27]. An MIT study found 95% of firms adopting AI have not yet seen any ROI on those investments ^[28]. Many analysts label high-flyers like Palantir – up 135% this year – overvalued, predicting a potential drop from its extreme $160+ share price to more “fundamental” levels (~$120) as initial euphoria wears off ^[29] ^[30]. Skeptics note that price-to-sales and P/E multiples for some AI leaders now rival dot-com-era extremes, fueling bubble comparisons.
Broader Shifts Boost and Threaten AI Rally: Developments beyond earnings are directly affecting AI stocks. Government policy is a big wild card: the U.S. is preparing new rules to curb reliance on foreign chipmakers, potentially mandating a 1:1 ratio of domestically-made to imported chips ^[31]. Companies that don’t meet the quota could face steep tariffs ^[32] – an America-first move that, if implemented, may reshape global supply chains. This could raise costs for Big Tech firms dependent on overseas fabs, yet benefit U.S. semiconductor players like Micron and Intel. In fact, Micron’s stock has been on a tear (up 43% in September) as it stands to gain from onshoring and surging memory demand ^[33] ^[34]. Demand for high-bandwidth memory (HBM) chips – critical for AI training – is so red-hot that Micron has sold out all its HBM production for 2025 already ^[35]. On the regulatory front, there’s also chatter of potential export restrictions (e.g. bans on Nvidia’s AI chips to certain countries), and in Europe, new AI regulations loom – factors that could cap some revenues. Meanwhile, macroeconomics are starting to pinch: rising long-term Treasury yields mean higher borrowing costs just as AI firms plan massive data center builds. As one analyst noted, the entire U.S. stock market’s fortunes now rest increasingly on hopes for AI growth – but higher yields could “complicate this picture” by making debt-funded AI investments less attractive ^[36].
Investor Sentiment and Outlook: Despite near-term wobbliness, the overall sentiment on Wall Street toward AI remains cautiously optimistic. The consensus is that AI is a real paradigm shift – but also a crowded trade. Earnings growth is starting to catch up: chipmakers like Micron reported record sales thanks to AI demand ^[37], and AI-exposed Big Tech (from cloud providers to chip suppliers) delivered strong recent quarters. However, stock prices have arguably priced in years of perfection. Many fund managers are rotating some profits out of the 2023 AI winners into more under-the-radar names that could lead “AI Boom 2.0” ^[38]. For instance, Broadcom and AMD – which provide networking gear and alternative AI chips – are touted as secondary beneficiaries of the spending wave ^[39] ^[40]. And new public entrants like CoreWeave could offer fresh ways to invest in the AI infrastructure build-out. Going forward, key catalysts will be actual product announcements and adoption metrics: e.g. the rollout of Nvidia’s next-gen chips, enterprise uptake of AI software like Microsoft’s Copilot, and government AI projects (the Stargate initiative, a proposed $500 billion AI datacenter network announced in January ^[41], is one to watch). Any evidence that AI is driving tangible productivity gains – or conversely, signs of corporate AI spending fatigue – will swing these stocks. In the near term, traders are watching for a possible “healthy correction” in overheated AI names ^[42]. But longer-term, the bulls argue that trillions in economic value are up for grabs, and today’s investments could herald the next tech revolution. As the dust settles on this week’s volatile trading, one thing is clear: the AI stock story is far from over, with new chapters of growth – or reckoning – still ahead.

Sources

Zacks/Nasdaq Market News – AI rally and stock movements (Sep 25, 2025) ^[43] ^[44]
Reuters – Nvidia’s $100B OpenAI deal and industry tie-ups ^[45] ^[46]; Oracle’s AI cloud contracts ^[47] ^[48]; U.S. chip policy plans (WSJ report) ^[49]
The Register – Oracle’s $18B debt raise & $455B backlog amid AI push ^[50] ^[51]
AInvest – AMD’s AI partnerships and valuation concerns ^[52]; Palantir’s surge and analyst warnings ^[53] ^[54]
Reuters Breakingviews – “AI investment bubble” commentary & MIT study ^[55] ^[56]
Reuters (Joachim Klement) – High yields risking the AI boom ^[57]
TipRanks (Barclays analysis) – Projected AI spending and Nvidia outlook ^[58] ^[59]
Finviz/InsiderMonkey – Analysts on Nvidia’s OpenAI partnership ^[60]
Nasdaq/Motley Fool – AI deals (Oracle, Microsoft, Nvidia) and Micron’s HBM demand ^[61] ^[62]