AI Stock Frenzy: July 2025’s Biggest Earnings Surprises, Deals, and Market Shake-Ups

Chipmakers Shine on Surging AI Demand

Major AI chipmakers saw dramatic moves in July 2025 amid booming demand and favorable policy news. Nvidia (NVDA) – now the world’s most valuable semiconductor firm – got a boost mid-month when the U.S. government reversed an export ban on advanced AI chips to China as part of a rare earth trade deal reuters.com reuters.com. Nvidia said it is applying for licenses to resume selling its H20 GPUs to Chinese customers and has been assured approvals are coming reuters.com. This marked a U-turn from an April ban that Nvidia warned would cost it up to $15 billion in lost revenue reuters.com reuters.com. Nvidia’s stock jumped about 4% on the news, as Chinese tech firms reportedly rushed to order the H20 chips – a slightly curbed version of its flagship AI processor – now back on the menu reuters.com reuters.com. Nvidia CEO Jensen Huang, visiting Beijing, argued that denying China access risked eroding Nvidia’s global leadership, noting the H20 works with Nvidia’s standard AI software tools even if its peak performance is dialed down reuters.com reuters.com.

Advanced Micro Devices (AMD) – Nvidia’s chief rival – likewise surged on the export reprieve. AMD confirmed the Commerce Department will review its applications to ship its MI308 AI accelerators to China and said it plans to restart those exports once licenses are approved reuters.com. AMD shares spiked as much as 7% on the announcement reuters.com. The export relief comes as AMD was already gaining momentum with new products. In June, AMD launched its Instinct MI350 series AI chips, which deliver four times the performance of prior models and are “on par with Nvidia’s latest-gen” GPUs, according to analysts ainvest.com ainvest.com. This prompted HSBC to double its price target on AMD from $100 to $200 on July 10, upgrading the stock to “Buy” and citing the MI350 as a credible challenger to Nvidia’s dominance ainvest.com ainvest.com. HSBC’s analysts were impressed that AMD’s new MI350 “can compete with Nvidia’s Blackwell GPUs,” and they see even more upside when AMD launches its next-gen MI400 chips in 2026, expected to rival Nvidia’s upcoming “Vera Rubin” platform investopedia.com. Other Wall Street firms echoed the optimism: on July 16, Wells Fargo raised its AMD target from $120 to $185 while affirming an “Overweight” rating, pointing to stronger-than-expected data center GPU sales and new high-end AI chip offerings like the MI355X ainvest.com ainvest.com. Bank of America also hiked its AMD target to $175 after estimating the company could ship $400–$600 million per quarter in AI GPUs in late 2025 – adding roughly $1 billion to 2025 revenue and $2 billion in 2026 as enterprises snap up AMD’s accelerators ainvest.com. Analysts note AMD is trading at a high valuation (~114× earnings), but its growing foothold in data-center AI and a slew of partnerships (with OpenAI, Meta, Oracle, Microsoft and others to use its forthcoming MI400 chips) underpin bullish long-term bets ainvest.com ainvest.com. As Morgan Stanley summed up, U.S. AI chip leaders are poised to capitalize on looser export rules: “US AI companies NVIDIA, AMD, and Broadcom are likely to receive licenses to ship to China at levels consistent with the prior threshold… it’s a significant positive for 2026 for all AI stocks, including our Top Pick NVDA” itiger.com. Still, Morgan Stanley urged investors to keep near-term expectations in check even as it reiterated Nvidia as a “top idea” for the AI boom itiger.com.

Big Tech Earnings: Cloud Titans Tout AI Growth

The late-July tech earnings season put AI front and center for U.S. cloud giants. Microsoft (MSFT), Alphabet (GOOGL), and Amazon (AMZN) all geared up to report results with investors laser-focused on their AI initiatives. Microsoft’s fiscal Q4 earnings (due July 30) were widely expected to show continued “explosive AI and cloud growth,” as one market observer put it swingtradebot.com. The company has woven AI into virtually all product lines – from Azure’s AI services to GitHub Copilot and Microsoft 365’s new AI features – driving optimism for strong results. Analysts broadly anticipate Microsoft’s momentum will continue unabated, riding the wave of enterprise AI adoption swingtradebot.com. Over at Google, parent Alphabet reported on July 23 with Wall Street keen to see how AI investments are translating into revenue swingtradebot.com swingtradebot.com. Google has been aggressively rolling out generative AI across its products – integrating its Gemini AI model into Google Workspace and revamping search with AI snapshots and interactive modes. In its Q1 call, Google disclosed that AI-driven features helped boost Google Cloud revenue 28% year-over-year swingtradebot.com. However, questions remain about monetization: Google’s AI search “Overviews” currently monetize at similar rates as regular queries, implying room for improvement as the company refines its strategy swingtradebot.com. Google is also navigating challenges from OpenAI and other upstart rivals in AI, even as it touts its own models like Bard and Gemini.

Amazon, meanwhile, gave a preview of the AI windfall in store. CEO Andy Jassy highlighted on Amazon’s last earnings call that the company’s AI business already has “a multibillion dollar annual revenue run rate, continues to grow triple-digit year-over-year…, and is still in its very early days” swingtradebot.com. He noted that prior to this generative AI era, Amazon Web Services (AWS) was on track to be a “multi-$100 billion run rate” business, and with generative AI the opportunity is even larger swingtradebot.com. Investors are eager for details on Amazon’s plan to capture that growth when it reports on July 31. Notably, Amazon has been developing its own AI chips (AWS Trainium and Inferentia) and large language models, aiming to reduce reliance on Nvidia and attract cloud customers with cost-effective AI infrastructure. Similarly, Meta Platforms (META) has been investing heavily in AI for content recommendations and the metaverse; in July it even poached several senior Google AI researchers as it vies to build cutting-edge AI models swingtradebot.com swingtradebot.com. And Apple (AAPL) – which faces criticism for lagging in AI – reportedly has its own “AI troubles” to overcome swingtradebot.com. All told, AI was the common theme on Big Tech earnings calls, with executives set to be grilled on how AI spend is translating into new revenue streams. As one headline quipped, Wall Street’s two favorite letters – “AI” – “once again dominate the conversation” this earnings season swingtradebot.com.

Software & Platform Stocks Hit Record Highs

Beyond the mega-cap hardware makers, enterprise AI software companies enjoyed skyrocketing stock prices in July. Data analytics and AI platform provider Palantir Technologies (PLTR) has been one of 2025’s standout winners. Palantir’s shares have more than doubled year-to-date (up 100% as of mid-July) and quadrupled since early 2024, recently hitting all-time highs above $150 tipranks.com. This surge – over 400% in about 18 months – reflects surging optimism around Palantir’s AI offerings and newfound profitability. In mid-July, Palantir announced it is partnering on a Pentagon project dubbed “Warp Speed for Warships,” aimed at using AI to accelerate U.S. Navy shipbuilding tipranks.com. The initiative adds to Palantir’s string of high-profile government deals and showcases its new Artificial Intelligence Platform (AIP), which helps large organizations embed ChatGPT-style intelligence into their operations nerdwallet.com tipranks.com. Palantir’s CEO Alex Karp has positioned the company at the forefront of military and enterprise AI adoption, and investors are paying attention. Even after the huge rally, some bulls remain optimistic – one top-ranked investor noted Palantir’s “unmatched momentum, government ties, and accelerating growth” as justification for continued upside, calling the company an integral part of the U.S. defense AI ecosystem tipranks.com tipranks.com. However, with the stock’s valuation now in the stratosphere, there are also calls for caution. The same investor warned that Palantir’s “nose-bleed valuation” means even a small earnings miss could trigger a sharp pullback, though he is sticking with a Buy rating for now tipranks.com tipranks.com. Wall Street consensus has also turned more neutral on Palantir after the torrid run – as of July the stock carries a majority of Hold ratings and even some Sells, with analysts’ average price target ($105) sitting well below the current trading price tipranks.com. All eyes will be on Palantir’s next earnings (expected in early August) to see if its results can live up to the AI hype.

Another pure-play enterprise AI name, C3.ai, Inc. (AI), also saw a resurgence. C3.ai’s stock jumped roughly 18% in a single month through mid-July, far outpacing the S&P 500’s ~4% gain in that period nasdaq.com. The move put C3.ai back on investors’ radar after a volatile few years since its 2020 IPO. The rally appears driven by renewed enthusiasm for companies offering turnkey AI solutions to businesses. C3.ai specializes in pre-built AI applications for things like supply-chain optimization, fraud detection, and predictive maintenance, rather than just AI infrastructure nasdaq.com. The company has also aggressively expanded partnerships – integrating its software with the cloud platforms of Microsoft, Amazon, and Google, and aligning with consultancies like PwC and McKinsey to help deploy its AI solutions in more industries nasdaq.com. These alliances give C3.ai broader market access via established sales channels and enterprise relationships. Importantly, C3.ai has staked an early lead in the emerging field of “agentic AI.” While many rivals only recently started discussing agent-based AI, C3.ai secured a patent in 2022 and already has 100+ agentic AI applications in production across defense, manufacturing, and government clients nasdaq.com. The company says its agentic AI products (which can autonomously make decisions in complex workflows) are generating about $60 million in annual recurring revenue – a figure some analysts believe underscores the stock’s undervaluation given the growth potential nasdaq.com. High-profile customers lend credibility: C3.ai’s solutions are used by the U.S. Air Force, energy giant Shell, pharma leader GSK, and even the State of California, indicating real-world traction beyond just hype nasdaq.com. After plunging 85% from its post-IPO peak, C3.ai’s recent rebound has prompted debate over whether it’s truly turning a corner or simply riding speculative fervor fool.com nasdaq.com. The stock remains about 37% below its 52-week high (it closed around $28.50 on July 22 vs. a high of $45) nasdaq.com, suggesting room to run if the company delivers consistent growth. But like many AI plays, C3.ai is not yet profitable, so investors are closely watching its quarterly results and guidance for confirmation that the accelerating demand for AI solutions will flow through to its bottom line.

Crucially, it’s not just Palantir and C3.ai – dozens of smaller AI-linked stocks have been soaring, reflecting a broader speculative wave. According to one index of AI and big data companies, the top performers over the past year include obscure names with astronomical gains. For example, tiny Quantum Computing Inc. (QUBT) has rocketed over 3,200% in one year, and SoundHound AI (SOUN) – a voice AI firm – is up about 126% nerdwallet.com nerdwallet.com. Even more established players like Upstart Holdings (UPST) (which uses AI for lending) have more than tripled, gaining over 200% in a year nerdwallet.com nerdwallet.com. These outsized moves underscore the intense risk appetite for anything AI-related. However, they also raise eyebrows about the sustainability of such rallies if fundamentals don’t catch up.

Deals, Launches and Industry Moves

In addition to earnings, July saw a flurry of partnerships and product unveilings in the AI sector. We’ve noted Palantir’s new defense contract and C3.ai’s expanded cloud alliances. Meanwhile, chipmakers and big tech pursued their own strategic moves. AMD’s launch of the MI350 GPU line (in June) has been followed by industry buzz over its upcoming MI400 series, which AMD is co-developing with key customers. Reports suggest AMD’s next-gen chips are being designed in collaboration with OpenAI, Meta, Oracle, Microsoft and others – a coalition aimed at ensuring Nvidia’s GPUs don’t remain the only game in town ainvest.com. This kind of industry partnership could give AMD a serious foothold if its 2026 MI400 launch delivers on promised performance gains. Intel (INTC), though not enjoying the same stock pop as Nvidia or AMD, also rolled out new AI accelerators and announced it is integrating AI capabilities into its popular Core PC processors shipping later this year. IBM made news by leveraging its watsonx AI platform in cloud deals, and Salesforce launched a suite of AI features for its software, including a $500 million fund to invest in AI startups (signaling the ongoing M&A and venture activity around AI). Even Tesla hinted at progress on its Dojo AI training supercomputer, which Elon Musk claims could eventually be offered as a cloud service. And in an interesting talent maneuver, Meta Platforms hired away a team of Google’s top AI researchers (including those behind Google’s award-winning AI model for protein folding) as the battle for AI expertise intensifies swingtradebot.com. While July didn’t see mega-mergers in the AI space, the continuous stream of smaller acquisitions and collaborations – whether it’s cloud providers teaming with model developers, or consultancies buying AI startups to bolster their practices – underscores that the AI ecosystem is rapidly evolving via partnership and consolidation. Industry insiders say we can expect more tie-ups in coming months as established players look to either acquire promising AI upstarts or at least secure exclusive alliances with them before competitors do.

Regulatory Tailwinds and U.S. AI Policy

In Washington, policymakers took significant steps in July that could shape the future of AI businesses. On July 16, the Biden administration’s prior hardline stance on AI chip exports was partially rolled back – as noted earlier, President Trump (in office since January) negotiated a deal to allow Nvidia and AMD to resume sales of advanced AI GPUs to China reuters.com reuters.com. This move, aimed at securing a trade agreement on rare earth minerals, drew mixed reactions. U.S. national security hawks raised concerns that granting China access to high-end AI chips – even slightly neutered versions – could aid Beijing’s military and AI firms like the mysterious startup DeepSeek (which claims startling AI efficiency) reuters.com reuters.com. Lawmakers from both parties questioned the reversal, noting the H20 chip had been identified as fueling China’s AI progress, and vowed to seek clarification on the strategic rationale reuters.com reuters.com. Nevertheless, the Commerce Department appears set to grant licenses, reflecting the administration’s calculation that U.S. companies need access to China’s massive market to sustain their edge. Nvidia’s Huang argued that if American firms can’t sell to Chinese tech giants, local alternatives (e.g. Huawei’s developing AI chips) will fill the void – potentially eroding U.S. dominance long-term reuters.com.

In a broader stroke, the White House unveiled a sweeping national AI strategy on July 23 designed to supercharge the U.S. AI industry. The “AI Action Plan” announced by President Trump represents a sharp pivot from the previous administration’s approach. It aims to “loosen environmental rules and vastly expand AI exports to allies” in order to maintain America’s edge over China reuters.com. The plan’s 90 recommendations include aggressively promoting the export of U.S. AI software and hardware to friendly nations, and cracking down on state-level regulations deemed too restrictive on AI development reuters.com. This is a marked departure from ex-President Biden’s “high fence” strategy that tightly controlled access to advanced AI chips abroad reuters.com. Michael Kratsios, the White House science & tech chief, said the government will partner with industry to deliver “secure full-stack AI export packages” – including chips, AI models, software and standards – to America’s allies worldwide reuters.com. Such a move could directly benefit U.S. tech giants: An expansion of exports “of a full suite of AI products” would help Nvidia and AMD (chips) as well as AI cloud and model providers like Google, Microsoft, OpenAI, and Meta, the plan notes reuters.com. In tandem, Trump rescinded several Biden-era orders that tech firms had opposed – including an AI “diffusion rule” that capped how much computing power certain countries could acquire via U.S. chips, and directives meant to police AI for misinformation or anti-competitive behavior reuters.com. The new stance was trumpeted by Vice President J.D. Vance, who warned against “[regulating] ourselves to death” while China catches up, calling some prior policies “stupid” for hindering American innovation reuters.com. Notably, the plan does not resolve remaining security debates such as whether to re-impose limits on Nvidia’s H20 chip – it explicitly sidestepped the controversy over that chip, which Trump had banned in April but then allowed in July, sparking rare GOP criticism of the pro-business move reuters.com. Beyond exports, the administration is also tackling infrastructure bottlenecks: the AI plan will fast-track data center construction by easing environmental and permitting rules, even offering use of federal lands for new server farms. This could alleviate one constraint on the AI boom – the shortage of data center capacity and electricity for training giant AI models – and further boosts companies building AI supercomputers. Overall, July’s policy shifts show Washington racing to enable an AI build-out: the government is essentially removing regulatory hurdles and encouraging U.S. firms to scale AI as quickly as possible, both domestically and in friendly overseas markets. This pro-industry stance was welcomed by many in Silicon Valley, though some consumer advocates worry it could lead to laxer oversight on AI safety and privacy. For investors, the takeaway is that U.S. policy is now firmly aligned with fueling the AI boom, which may extend the runway for AI-related stocks – albeit with the caveat that geopolitical tensions remain a wildcard (export policies could tighten again if national security fears outweigh economic aims).

AI ETFs Boom as Analysts Debate “Bubble” vs. Reality

With AI the hottest theme on Wall Street in 2025, AI-focused funds and indices continued to surge in July. The frenzy of inflows into AI-themed ETFs has been remarkable – by some estimates, U.S. thematic tech ETFs (dominated by AI and robotics funds) have seen record inflows and performance. For instance, the Global X Artificial Intelligence & Technology ETF (AIQ) and the Roundhill Generative AI ETF (CHAT) are both significantly outperforming the broader market this year (each up well over 50% year-to-date). The Indxx Global Robotics & AI Index, which tracks dozens of AI-exposed stocks, is boasting eye-popping trailing returns thanks to the kind of high-flyers noted above. According to NerdWallet, the seven best-performing AI stocks in that index have all at least doubled in the past year – including Quantum Computing Inc.’s +3,219% explosion and Palantir’s +425% gain nerdwallet.com nerdwallet.com. Even larger constituents like AppLovin (APP) (mobile app AI) and Upstart have notched triple-digit percentage increases nerdwallet.com nerdwallet.com. These broad-based rallies have lifted AI ETFs and mutual funds, some of which are hitting new all-time highs. By late July, AI mania has visibly helped propel the overall stock market higher – the tech-heavy Nasdaq Composite is near 18-month highs, and the S&P 500’s year-to-date gains are largely driven by a handful of AI-centric mega-caps.

However, the central question nags at many: is this the start of a sustained technological revolution that justifies soaring valuations, or an overheated bubble echoing the late-90s dot-com era? In a thoughtful analysis on July 22, Reuters compared today’s AI boom to the 1999–2000 tech bubble. There are certainly frothy signals: As Apollo Global Management’s chief economist Torsten Slok pointed out, the top 10 stocks in the S&P 500 now trade at higher forward earnings multiples than at the peak of the dot-com bubble reuters.com. The market cap concentration is also extreme – those 10 companies (dominated by Big Tech) comprise about 40% of the S&P’s value today, versus 25% in 1999 reuters.com reuters.com. On the other hand, many of those firms are vastly more profitable and entrenched than the dot-com highfliers, and overall market valuations (e.g. the Nasdaq at ~28× forward earnings) are still below 2000 levels reuters.com. Crucially, unlike the dot-com era which was fueled by eyeballs and optimism, the AI wave is underpinned by tangible breakthroughs in computing power and real enterprise use cases – albeit expectations are sky-high. Morgan Stanley estimates that realizing the full potential of generative AI will require about $2.9 trillion in global data center investment by 2028 (for chips, servers, and infrastructure) reuters.com. Such a massive capital cycle could indeed drive years of growth for AI suppliers, but it also implies huge spending with uncertain near-term payoffs reuters.com reuters.com. As the Reuters piece noted, a “meaningful, prolonged market correction cannot be ruled out” if AI-driven growth “isn’t delivered as quickly as investors expect.” reuters.com reuters.com For instance, if advancements like autonomous robots, fully AI-driven customer service, or drug discovery AI take longer than optimists project, some of today’s lofty sales forecasts may need to be reined in. Even leaders like Nvidia acknowledge that supply constraints (like limited 3nm chip fabrication capacity) and new competition (startups or in-house chips from companies like Amazon and Google) could eventually cap their growth swingtradebot.com.

Several veteran investors have been preaching vigilance. In one CNBC interview, hedge fund manager Dan Niles – formerly a noted tech bear – revealed he has turned bullish on Nvidia, precisely because the company’s recent actions and results convinced him that AI demand is real and still ramping. Nvidia took a large revenue write-down anticipating the China export ban, which derisked its outlook, and Niles observed that while AI training demand may be normalizing, inference demand (real-world usage of AI models) is now exploding swingtradebot.com swingtradebot.com. He cited data that in May, Google’s generative AI services produced 50× more output tokens than a year ago, and Microsoft’s Azure OpenAI usage jumped 5× – concrete indicators that end-user adoption of AI (chatbots, assistants, etc.) is surging swingtradebot.com. This shift from hype to actual usage gives credence to the idea that AI could deliver sustained revenue streams. Niles also pointed out that sovereign wealth funds and international governments are pouring money into AI (President Trump’s visits to the Middle East yielded multi-billion-dollar AI partnership deals, for example), creating additional backstop demand for U.S. AI firms swingtradebot.com. Nonetheless, even Niles cautioned that Nvidia won’t have the field to itself indefinitely – companies like Apple and Qualcomm are vying for the same advanced chip manufacturing capacity, and cloud players like Amazon are developing in-house chips (e.g. AWS’s Trainium2) that promise cost savings and could divert some workloads away from Nvidia swingtradebot.com.

In summary, July 2025 was a milestone month in the AI stock saga – featuring blockbuster rallies, bold corporate bets, and supportive government moves. The AI revolution is clearly driving real growth for many companies, from chipmakers enjoying unprecedented order backlogs to software firms signing huge new contracts. Yet, the exuberance around anything AI-related has also become intense. Seasoned voices remind us that execution and innovation must justify the valuations in the long run. For now, the market’s love affair with AI shows few signs of cooling. As one investment fund wrote in its quarterly letter, “With the AI market still in the early stages of a multi-year infrastructure build-out, NVIDIA is well-positioned to capture significant value as the industry standard” swingtradebot.com – a sentiment that could apply to many leaders in this space. Investors will be watching closely in the coming months for confirmation that this AI gold rush can translate into sustained earnings growth. If July was any indication, the second half of 2025 promises to be an exciting – and possibly volatile – ride for all things AI in the stock market.

Sources: Earnings and analyst commentary from Yahoo Finance, Reuters, and Bloomberg; company press releases; Reuters analysis on tech valuations reuters.com reuters.com; Reuters reports on U.S.–China chip export policy reuters.com reuters.com and White House AI initiatives reuters.com reuters.com; industry insights from TipRanks and Zacks on Palantir and C3.ai tipranks.com nasdaq.com; NerdWallet data on top-performing AI stocks nerdwallet.com nerdwallet.com; and CNBC interviews with market experts swingtradebot.com swingtradebot.com.