LIM Center, Aleje Jerozolimskie 65/79, 00-697 Warsaw, Poland
+48 (22) 364 58 00
ts@ts2.pl

AI Stock Frenzy: Google's $210 Billion Win, Nvidia's Trillion-Dollar Boom, Microsoft's Mega AI Deal & Tesla's Bold Pivot in 48 Hours

AI Stock Frenzy: Google’s $210 Billion Win, Nvidia’s Trillion-Dollar Boom, Microsoft’s Mega AI Deal & Tesla’s Bold Pivot in 48 Hours
  • Alphabet (Google) stock surged nearly 9% (adding $210 billion in value) after a court ruled against breaking up Google, citing rising AI competition reuters.com reuters.com.
  • Nvidia signaled record AI demand – CEO Jensen Huang insists “everything [is] sold out” as he forecasts a $3–$4 trillion AI infrastructure market by 2030 reuters.com reuters.com.
  • AMD executives dismiss “AI bubble” fears, projecting a $500 billion+ AI chip market by 2028 and ramping up next-gen AI GPUs despite export curbs wccftech.com investing.com.
  • Microsoft struck a $3 billion AI partnership with the U.S. government, offering free Copilot AI tools to millions of federal users for a year to speed up AI adoption blogs.microsoft.com capacitymedia.com.
  • Palantir shares jumped midweek as enthusiasm for its AI Platform and steady government contracts drove momentum hotcandlestick.com; Wall Street analysts hiked price targets (Wedbush now $200, +25%) on robust demand nasdaq.com.
  • Tesla shuttered its “Dojo” AI supercomputer project, with Elon Musk redirecting resources to Tesla’s next-gen AI5/AI6 inference chips reuters.com reuters.com – a dramatic pivot given Dojo’s earlier ~$500 billion valuation by Morgan Stanley reuters.com.
  • AI earnings shake-up: Enterprise AI firm C3.ai reported a 19% revenue drop and 86% wider loss, called “completely unacceptable” by its founder ainvest.com. The company replaced its CEO and withdrew guidance, sparking a 25% stock slide this month ainvest.com ainvest.com.

Alphabet Soars on Antitrust Victory Fueled by AI

Alphabet Inc. – Google’s parent – saw its stock rocket 9% to a record high on Sept 3 after a U.S. judge refused to break up the company in an antitrust case reuters.com. The ruling lifted a huge regulatory cloud and explicitly cited the rise of AI tools like ChatGPT as emerging search competition undermining monopoly concerns reuters.com. Judge Amit Mehta opted for targeted remedies (like data-sharing with rivals) over drastic measures, a decision analysts called “pragmatic” and a clear win for Google reuters.com reuters.com. The outcome preserves Google’s control of Android and Chrome and its lucrative search deals – allowing it to deepen partnerships like integrating its upcoming “Gemini” AI into Apple’s products reuters.com. “This removes a significant legal overhang,” said Hargreaves Lansdown analyst Matt Britzman, noting the court’s willingness to pursue practical fixes over “scorched-earth tactics” reuters.com. Alphabet’s market cap swelled by over $200 billion in a day reuters.com, buoying the entire tech sector (Apple shares jumped ~4% on relief that Google can keep paying for Safari search default reuters.com). The ruling’s acknowledgment of AI-driven competition validated Google’s massive AI investments and eased fears of structural breakup, electrifying investors.

Chipmakers: Nvidia, AMD (and Broadcom) Ride the AI Wave

Chip stocks remained front-and-center as Nvidia – the world’s AI silicon leader – signaled that the AI boom is far from over. Days after reporting blowout earnings, Nvidia’s CEO Jensen Huang projected $3–$4 trillion in AI infrastructure spending by 2030 reuters.com. “A new industrial revolution has started. The AI race is on,” Huang declared, emphasizing that demand for Nvidia’s AI GPUs remains insatiable. “The buzz is: everything sold out,” he said of its H100 and new H200 processors reuters.com. Nvidia’s stock, which has soared over 10-fold since 2023, paused its rally in early September amid some profit-taking and a “tepid” new sales forecast that excluded potential China revenues reuters.com. (Nvidia is navigating U.S.-China export curbs – its forecast assumed zero sales to China, underscoring ongoing trade uncertainty reuters.com.) Still, analysts see no signs of demand slowing. “If anything, this just highlights the durability of this AI trade,” said Raymond James’s Matt Orton, noting no slowdown in Nvidia’s results reuters.com. Indeed, Nvidia’s Q2 revenue surged 56% year-on-year to $46.7 billion nvidianews.nvidia.com, and it issued an ambitious Q3 outlook of $54 billion (even without China) nvidianews.nvidia.com. The slight stock dip (~2% post-earnings) is seen as a breather in an otherwise historic run reuters.com – Nvidia’s valuation remains elevated, but many on Wall Street argue the AI spending cycle is only in its early innings.

Advanced Micro Devices (AMD), Nvidia’s chief rival, also grabbed headlines by doubling down on its AI ambitions. At Citi’s Global Technology Conference on Sept 3, AMD executives rejected talk of an “AI bubble” and painted AI as a “once-in-a-lifetime opportunity” wccftech.com. CFO Jean Hu noted “we are still at a very early stage for AI adoption,” pointing to “tremendous evidence for AI” uptake across industries wccftech.com. AMD expects the market for AI accelerators to exceed $500 billion by 2028 wccftech.com, and it’s racing to carve out a bigger share. The company highlighted a robust roadmap of AI chips: its current flagship MI300 GPU is ramping up, with next-generation MI400 chips planned for 2025 and MI500 in 2027 investing.com. AMD is aggressively partnering with major cloud players and even sovereign governments (over 40 nations) to deploy its AI solutions investing.com. However, AMD also acknowledged challenges – notably U.S. export controls to China that led to a hefty $800 million inventory write-off for its China-specific MI308 chips investing.com. Still, AMD’s tone was confident. It sees AI as “the biggest inflection in computing since the invention of the Internet,” according to AMD’s Matt Ramsay, and is positioning to challenge Nvidia’s dominance with a broad suite of data center and inference products wccftech.com. Investors have taken note: AMD shares, up over 70% year-to-date, have been volatile but reflect optimism that it can capture a growing slice of the AI silicon pie.

Broadcom, another key player, was poised for its own AI moment. The semiconductor giant (which supplies networking chips and custom AI accelerators) was set to report earnings on Sept 4, with investors eyeing its surging AI revenue. In the prior quarter, Broadcom’s sales from AI-related semiconductors hit $4.4 billion (up 46% YoY), and the company forecast $5.1 billion in AI chip revenue for the August quarter – nearly one-third of total sales fool.com. Broadcom’s overall revenue was expected to jump ~20%, fueled largely by hyperscale cloud demand for AI fool.com. Analysts predicted a solid beat, noting that Broadcom’s $15.8 billion Q3 revenue guidance already reflected sizable AI traction finance.yahoo.com. The stock had rallied into the results (up ~1.4% to $302 on Sept 3 stockinvest.us), with bullish sentiment that Broadcom’s unique position in supplying cloud AI infrastructure (from networking gear to Google/Tesla partnership chips) would “bellwether” the AI trade nai500.com. As one of the last big-tech earnings of the season, Broadcom’s report was seen as a key barometer for whether the “AI hype” translates into tangible sales across the sector. Early indications were positive – the company has emphasized that AI will be a long-term growth driver, not a one-quarter wonder nai500.com. (Indeed, its pending VMware acquisition is partly to bolster AI-enabled software offerings nai500.com.) Broadcom’s results were due after market close on the 4th, with any surprises likely to reverberate through chip stocks in the days after.

Microsoft Lands a $3 Billion AI Deal with Uncle Sam

Microsoft made waves by inking a groundbreaking agreement with the U.S. government aimed at accelerating AI adoption across federal agencies. Announced on Sept 2 and cheered by investors through the week, the deal – brokered with the General Services Administration – will provide millions of government workers with access to Microsoft’s latest AI and cloud software. Notably, Microsoft is offering its 365 Copilot AI assistant free for up to 12 months to all existing “G5”-tier agency users blogs.microsoft.com. This essentially gives federal employees a taste of generative AI tools (like GPT-powered Office apps, chatbots, and automation) at no cost for a year, jump-starting usage. In addition, Microsoft is throwing in steep Azure cloud discounts and waiving data transfer fees to help agencies modernize with AI blogs.microsoft.com capacitymedia.com. The initial one-year value is estimated at $3 billion in cost savings to the government blogs.microsoft.com, and agencies can lock in discounted pricing for up to three years. Microsoft even committed $20 million in support services to ensure successful implementation capacitymedia.com – underlining how badly it wants this AI push to succeed.

This is a major strategic win for Microsoft’s enterprise AI business. It not only cements Microsoft as the primary AI partner for the U.S. government but also showcases confidence in the security of its AI offerings (Copilot has already cleared key DoD security reviews) blogs.microsoft.com. The deal aligns with the White House’s “AI Innovation” plan, effectively making Microsoft’s platform a default toolkit for civil servants – from automating call centers with AI agents to using Copilot for data analysis and workflow automation blogs.microsoft.com blogs.microsoft.com. Analysts lauded the move as both a reputational boost and a future revenue driver: once the free period ends, the government will likely scale up paid subscriptions given the productivity gains. Microsoft shares, already near all-time highs after a year of AI-fueled optimism, held steady on the news – a sign that the market sees this as reinforcing Microsoft’s entrenched position in AI cloud services. CEO Satya Nadella has often said “AI is the new runtime of Microsoft’s platform,” and this OneGov GSA pact firmly embeds that platform across the public sector. In short, Microsoft just turned the U.S. government into one of its largest AI customers – a long-term tailwind for its Azure and Office 365 franchises as the AI era unfolds.

Palantir’s Rally: Upgrades and Global AI Tie-Ups

Data analytics firm Palantir Technologies – often dubbed an “AI play” for its government-focused software – enjoyed a resurgence during the week, powered by both Wall Street and international partners. Palantir’s stock climbed on Sept 3 (one of its best days in weeks) as investors piled back in, encouraged by “enthusiasm around its AI platforms and [a] steady flow of government contracts,” according to Yahoo Finance hotcandlestick.com. The company has been touting its new Artificial Intelligence Platform (AIP), which helps enterprises deploy large language models on private data. That narrative gained credence with a notable partnership overseas: Palantir extended its alliance with Japan’s Fujitsu to roll out Palantir’s AIP to Japanese customers prnewswire.com prnewswire.com. Under a licensing deal signed in August, Fujitsu will integrate Palantir’s generative AI platform into its suite (Fujitsu Uvance) and offer it to clients across Japan, with plans to expand globally in 2025 prnewswire.com prnewswire.com. This effectively gives Palantir a major distribution partner in Asia and underscores the worldwide demand for its AI solutions in industries from finance to manufacturing.

Crucially, big-name analysts dramatically raised their targets on Palantir, adding fuel to the stock’s momentum. In late August, Citi and Piper Sandler each hiked their Palantir price targets into the high-$170s after stunning Q2 results nasdaq.com. And on Sept 3, Wedbush Securities – pointing to “strong demand” for Palantir’s AI offerings – boosted its target to $200 from $160, one of the highest on Wall Street nasdaq.com. (For context, Palantir traded around the mid-$150s this week, so $200 implies confidence in significant upside.) The wave of bullish calls follows Palantir’s announcement that U.S. government revenue jumped 53% and that it’s positioning itself as a key AI contractor for defense and intelligence agencies nasdaq.com. “Show-stopping” earnings and the launch of Palantir’s new AI platform have flipped sentiment, with many analysts now viewing Palantir as a prime beneficiary of the AI spending boom in both government and commercial arenas nasdaq.com.

That said, not everyone is drinking the Kool-Aid – noted short-seller Andrew Left of Citron Research cautioned that Palantir’s valuation is veering into “bubble” territory and recently compared its hype unfavorably to OpenAI nasdaq.com. Indeed, Palantir’s stock had pulled back about 18% in late August after a torrid run (it was still up ~100% in 2025) hotcandlestick.com. But this week’s news – analyst validation and evidence of global traction – helped Palantir regain its stride. The company’s charismatic CEO Alex Karp, who has called Palantir “the most important software company of our time” for AI in warfare and industry, now has fresh proof points to back that bold claim. Investors seem to agree, at least for now: Palantir is back on offense, leveraging its deep government ties and new AI tech to stake out a leading position in the big data AI gold rush.

Tesla Pivots: Dojo Supercomputer Shelved for In-Car AI Chips

Electric vehicle maker Tesla made a striking strategic U-turn in its AI efforts that had analysts buzzing. CEO Elon Musk has disbanded Tesla’s “Dojo” supercomputer team, effectively shutting down the ambitious in-house project to build a custom AI training supercomputer reuters.com reuters.com. This move – first reported by Bloomberg and confirmed by Musk on Aug 8 – was officially acknowledged going into September, marking an end to a high-profile initiative that Musk once touted as a potential game-changer (and which some on Wall Street valued astronomically). Musk explained the rationale bluntly: “It doesn’t make sense for Tesla to divide its resources and scale two quite different AI chip designs,” he wrote on X (Twitter) reuters.com. Instead, Tesla will focus all its chip engineering on its upcoming “AI5” and “AI6” chips – specialized inference processors to run AI models in real time inside Tesla cars and robots reuters.com. “All effort is focused on [our next-gen] chips,” Musk said, implying Tesla will rely on those for both self-driving software and its humanoid Optimus robots rather than pursuing a separate Dojo training system reuters.com reuters.com.

The about-face is significant. Dojo, first unveiled in 2021, was Tesla’s plan to use custom silicon and a massive supercomputer to train its Autopilot/FSD neural nets at ultra-high speed. Hype around Dojo was so great that Morgan Stanley once estimated it could add $500 billion to Tesla’s valuation by unlocking new AI services reuters.com. Now, Musk has essentially deemed it an “evolutionary dead-end.” About 20 Dojo engineers already left Tesla to form a startup, and the rest have been reassigned internally reuters.com. Rather than building a giant AI computing cluster itself, Tesla is outsourcing some of that to partners – notably, Musk inked a $16.5 billion deal with Samsung to produce Tesla’s next-gen AI6 self-driving chips reuters.com. These chips (due by 2026) will power Tesla’s vision for fully autonomous driving and robotics, performing both the perception and decision-making inference on-device. Musk claims the AI5/Ai6 chips “will be excellent for inference and at least pretty good for training” reuters.com, suggesting Tesla can get by without a separate Dojo for now.

Investors are digesting what this means. On one hand, shelving Dojo could reduce cash burn and R&D risk, and focusing on car AI chips is in line with Tesla’s core business (making cars smarter). Tesla has also had to prioritize amid various challenges – EV sales pressure, Musk’s other ventures, and multiple exec departures in the past year reuters.com. On the other hand, abandoning a moonshot AI project like Dojo raises questions: Did Tesla hit technological roadblocks? Will it fall behind others building AI supercomputers (like Google’s TPU farm or Nvidia’s DGX Cloud) for training? Musk’s decision indicates he’d rather lean on external AI infrastructure (possibly using more third-party chips or cloud services) and concentrate Tesla’s talent on the AI chips that go into its vehicles. Notably, Tesla’s stock didn’t react dramatically to the Dojo news – it’s up over 100% this year largely on EV optimism and the AI narrative, and some think this pragmatic move could be positive. “Dojo was sexy, but car sales pay the bills,” quipped one analyst. In any case, Tesla remains deeply committed to AI – but in a targeted way, aiming to deliver on the long-promised Full Self-Driving (FSD) capability and perhaps monetize AI in robots down the road, rather than chasing supercomputing glory for its own sake.

AI Earnings Spotlight: C3.ai’s Slump Triggers a Shake-Up

Not all AI stock stories were rosy. C3.ai, a prominent AI software company, delivered a sobering quarterly report that sent its shares sliding and forced a leadership overhaul. On Sept 3, C3.ai announced fiscal Q1 2026 results that fell well short of expectations: revenue plunged 19% year-over-year to about $70 million, while net losses ballooned 86% to $117 million ainvest.com ainvest.com. It was one of C3’s worst performances since going public. CEO Thomas Siebel – the billionaire founder of the company – didn’t mince words, calling the quarter “completely unacceptable” ainvest.com. He attributed the poor results to “weak sales execution” and internal disruptions, including mid-quarter management turnover and even his own health-related absence ainvest.com. In the earnings call, a chastened Siebel admitted C3 had execution problems and announced a major restructuring to refocus the sales organization.

The company took the dramatic step of appointing a new CEO, effective immediately. Stephen Ehikian (co-founder of enterprise software firm ThoughtSpot) was named chief executive, with Siebel shifting to an Executive Chairman role – a bid to restore investor confidence with fresh leadership. C3 also withdrew its prior full-year guidance, signaling that it lacks visibility and didn’t want to over-promise ainvest.com. Instead it issued only next quarter’s outlook, which was notably conservative (projecting at most $80 million in revenue and ongoing hefty losses) ainvest.com. All these red flags sent C3.ai’s stock tumbling: shares fell over 10% in after-hours trading, and as of Sept 4 the stock was down more than 25% from its late-August level ainvest.com ainvest.com.

The swift downturn marks a sharp reversal of fortune. C3.ai had been a retail trader favorite and one of the original “AI meme stocks” earlier in 2023, at one point soaring on the hype around artificial intelligence. The company sells enterprise AI applications (for example, AI-driven predictive maintenance for industries) and it rode the wave of enthusiasm to issue upbeat projections last year. But now reality is biting: growth has stalled and losses are widening, raising concerns about whether C3 was more sizzle than steak. In response, Siebel emphasized that C3.ai still has a strong technological foundation – boasting “131 enterprise AI applications” and a unique agent-based AI platform ainvest.com – and he pointed to a “$2 trillion” long-term AI market opportunity that the company could tap into. However, investors are in “show me” mode. The leadership shake-up is an attempt to get the company back on track. Siebel noted that newly appointed CEO Ehikian and other recent hires bring much-needed operational rigor, and the team is “ready to accelerate market penetration and move toward profitability” ainvest.com. Still, given the stock’s sharp sell-off, it’s clear the market is skeptical. C3.ai will need to prove it can convert AI buzz into real sales growth – otherwise, as this week showed, the AI hype can swiftly turn into a sell-off.

Sources:

The AI Bubble Is Popping (Which Will Make Smart Investors Rich)

Tags: , ,