11/30/2025
For your information.
ttps://www.msn.com/en-us/money/companies/michael-burry-exposes-how-big-tech-pumps-ai-profits/ar-AA1RoR7U?ocid=msedgdhp&pc=DCTS&cvid=6684a38c1b9c4774a1a4bca9d613efd3&ei=100
Michael Burry exposes how Big Tech pumps AI profits
Big Tech’s artificial intelligence boom is being sold as a revolution in productivity, but the numbers suggest it is also a carefully engineered financial machine. Michael Burry has zeroed in on how a handful of platforms are using AI narratives, accounting choices, and capital spending cycles to magnify reported profits and market power in ways that ordinary investors can easily misread.
I see his warnings less as a prediction of imminent collapse and more as a roadmap to where the incentives are skewed: in cloud infrastructure, in opaque AI “cost savings,” and in the way a few dominant firms can shift entire index returns. Understanding those mechanics is essential before treating AI as a one-way ticket to permanent earnings growth.
Michael Burry’s latest warning on AI-fueled profits
Michael Burry has built his reputation on spotting when financial stories drift too far from underlying cash flows, and his recent focus on AI fits that pattern. He has argued that the market is again concentrating risk in a narrow group of companies whose valuations depend on aggressive assumptions about future technology-driven profits, echoing the dynamic he previously highlighted in housing and in the 2021 meme-stock surge. In his view, AI has become a narrative that can justify almost any multiple as long as investors believe that current spending will translate into dominant, high-margin platforms later.
Related video: Alphabet Climbs Toward $4T As Nvidia Falls On AI Trade Shift (Benzinga)
Alphabet Climbs Toward $4T As Nvidia Falls On AI Trade Shift
His portfolio moves reflect that skepticism. Regulatory filings show that Burry has used index options and targeted shorts to bet against broad market benchmarks that are heavily weighted toward mega-cap technology names, while also taking selective long positions in more reasonably priced businesses that could benefit from AI without being priced for perfection. Those filings indicate that he is not dismissing AI as a technology, but rather questioning whether today’s earnings and cash flows support the valuations attached to the companies most loudly promoting it, a distinction that aligns with his earlier critiques of speculative manias in growth stocks and crypto-linked assets through filings and public comments.
How Big Tech turns AI spending into market power
The core of Burry’s concern is that AI is reinforcing an already extreme concentration of power in a few platforms that control cloud infrastructure, data, and distribution. Companies such as Microsoft, Alphabet, Amazon, and Meta are pouring tens of billions of dollars into data centers and specialized chips, then using that scale to lock in customers and partners. That spending is not just about building new products, it is also about raising the barrier to entry so high that smaller rivals cannot realistically compete on model training, inference speed, or global reach
Recent earnings reports show how this strategy works in practice. Microsoft has tied its Azure growth to AI services that are deeply integrated into Office 365 and GitHub, while Alphabet has bundled its Gemini models into search and Workspace, and Amazon has leaned on its AWS footprint to pitch AI tools to existing cloud clients. Each of these companies has reported that AI-related demand is a key driver of cloud revenue acceleration, and they have highlighted multi-year customer commitments that effectively cement their positions. Those disclosures, detailed in their latest earnings releases, Alphabet filings, and Amazon updates, show how AI capital spending is being converted into durable market share rather than just experimental R&D.
The accounting tricks behind AI “profitability”
On the surface, Big Tech’s AI push looks wildly expensive, yet reported margins have held up better than many expected. A key reason is accounting. Data centers, GPUs, and networking gear are treated as capital expenditures that are depreciated over several years, which means only a fraction of the actual cash outlay hits the income statement in any given quarter. That timing gap allows companies to present AI initiatives as accretive to earnings even while they are writing enormous checks to chipmakers and construction firms.
Several firms have also adjusted the useful lives of their servers and networking equipment, a change that lowers near-term depreciation expense and boosts operating income. Alphabet, for example, extended the estimated life of its servers and certain network gear, a move that added billions of dollars to operating profit over a full year according to its financial disclosures. Meta made similar changes to its depreciation schedules, which it said would improve 2024 operating income by several billion dollars in its quarterly report. Those choices are allowed under accounting rules, but they make AI-heavy businesses look more profitable today than they would if all costs were expensed immediately, a nuance that Burry has flagged as a source of investor overconfidence.
Cloud, chips, and the AI capex flywheel
AI’s financial engine runs through a tight loop connecting cloud providers, chipmakers, and software platforms. Big Tech companies commit to massive capital expenditure programs to build or lease data centers and secure priority access to advanced GPUs. Those orders, in turn, drive record revenue and pricing power for semiconductor firms, which then justify their own elevated valuations and expansion plans. The result is a feedback loop in which each side’s optimism about AI demand validates the other’s growth story, even if the ultimate end-user monetization is still uncertain.
Recent guidance from Nvidia, for instance, has highlighted extraordinary demand for its data center products, with revenue surging as hyperscalers race to deploy its H100 and successor chips. That demand is directly tied to the capex plans of Microsoft, Alphabet, Amazon, and Meta, which have all raised their full-year infrastructure spending forecasts in their latest financial updates and investor presentations. Burry’s concern is that this flywheel can obscure the underlying risk: if AI workloads do not generate the expected high-margin software and subscription revenue, the industry could be left with overbuilt capacity and hardware inventories that are far less productive than current models assume.
AI narratives and the new “Nifty Fifty” concentration
The AI boom has intensified an already sharp concentration of market returns in a small cluster of mega-cap stocks. Index performance in the United States has been dominated by a handful of technology and communication services names whose weightings have swelled as their share prices climbed. That concentration means that broad benchmarks can appear healthy even if the median stock is flat or declining, a pattern that Burry has previously criticized when he warned about narrow leadership in growth indices.
Data from recent index performance reports show that companies such as Microsoft, Apple, Alphabet, Amazon, Meta, and Nvidia account for a disproportionately large share of total market capitalization and earnings growth. Analysts have compared this group to the “Nifty Fifty” of the 1970s, a set of blue-chip stocks that investors once believed could be bought at any price because of their perceived invincibility. Today, AI is the unifying story that underpins similar confidence in the current leaders, as reflected in their premium valuation multiples and in the way passive flows automatically reinforce their dominance through capitalization-weighted index construction, a trend documented in index concentration studies and market structure analyses.
Where AI really boosts productivity, and where it doesn’t
One reason AI stories are so powerful is that there are genuine productivity gains in specific use cases, from code generation to customer support. Enterprise software vendors have reported that tools like AI copilots can reduce routine coding time, help draft documents, and automate parts of sales and marketing workflows. Those improvements can translate into higher willingness to pay for software subscriptions and can support price increases or new premium tiers, which in turn feed into Big Tech’s revenue growth.
The gap between promise and reality appears when those gains are extrapolated across entire economies or used to justify blanket assumptions about margin expansion. Surveys of corporate adopters show that while some teams see measurable efficiency improvements, others struggle with integration costs, data quality issues, and the need for human oversight. Several companies have disclosed that AI features are still a small fraction of total usage or revenue, even when they are heavily marketed in product launches and earnings calls. That nuance shows up in detailed breakdowns from firms like Microsoft and Salesforce, which have described early traction for AI add-ons but also noted that they remain in the ramp-up phase in their earnings commentary and investor materials. Burry’s skepticism rests on this mismatch between selective success stories and the sweeping claims often used to support market-wide AI valuations.
Regulatory scrutiny of AI dominance and data use
As AI becomes central to Big Tech’s business models, regulators have started to question whether the same companies that dominate search, social media, and cloud should also control the infrastructure and data pipelines for generative models. Antitrust authorities in the United States and Europe have opened inquiries into cloud market structure, data access, and the competitive impact of exclusive partnerships between large platforms and leading AI labs. Those investigations reflect a concern that incumbents could use their financial and technical advantages to entrench their positions in the next wave of computing.
Recent regulatory actions include formal probes into cloud pricing practices, reviews of major AI-related acquisitions, and scrutiny of how training data is collected and used. European regulators have examined whether bundling AI services with existing productivity suites could disadvantage smaller rivals, while U.S. agencies have signaled that they are watching large investments in independent AI developers for potential control or influence. These moves are documented in official competition announcements and regulatory inquiries. For investors, the risk is that business models built on aggressive data collection or tight ecosystem lock-in could face legal or structural constraints just as AI-driven revenue streams are being scaled up.
What Burry’s playbook suggests for AI investors
Looking at Burry’s history, his approach to AI-linked markets appears to follow a familiar pattern: focus on balance sheets and cash flows, question consensus narratives, and look for asymmetries where downside is underpriced. He has often emphasized that the most dangerous moments are when investors extrapolate recent trends indefinitely, whether in housing prices, passive flows, or now AI-driven earnings. In practical terms, that means stress-testing how sensitive Big Tech valuations are to slower AI adoption, higher capital costs, or regulatory friction.
For individual investors, his playbook points toward a few concrete disciplines. One is to separate the undeniable long-term potential of AI from the specific claims made by any single company about its competitive moat or margin trajectory. Another is to examine how much of a firm’s reported profitability depends on accounting choices such as extended asset lives or capitalized software development, details that are spelled out in the notes to financial statements and in regulatory filings. A third is to recognize concentration risk in portfolios that track major indices, where AI leaders can dominate performance. Burry’s own use of hedges and selective shorts, documented in his 13F disclosures, underscores his belief that even transformative technologies can be overpaid for when enthusiasm outruns verifiable earnings power.
The real test for AI profits is still ahead
For now, Big Tech’s AI push is delivering impressive headline numbers, from surging cloud revenue to record chip sales. The question Burry keeps pressing is whether those gains represent a sustainable new profit base or a phase in a capital cycle that could eventually normalize. The answer will depend on how quickly AI applications move from pilot projects and early adopters into the core workflows of industries like healthcare, manufacturing, and finance, and whether customers accept the pricing power that platforms are trying to exert.
Over the next few years, investors will be able to track that transition in a few tangible ways: the share of revenue tied explicitly to AI products, the trajectory of operating margins after the initial wave of depreciation adjustments, and the outcome of regulatory challenges to data and cloud dominance. Company disclosures already provide some of these signals in segment reporting and forward guidance, as seen in the detailed breakdowns from Microsoft, Alphabet, Amazon, Meta, and Nvidia in their latest earnings reports, investor updates, and financial statements. Burry’s critique serves as a reminder that the real measure of AI’s economic impact will not be the size of the data centers or the speed of the chips, but the durability of the cash flows that emerge once the initial investment frenzy subsides.
kommonsentsjane
kommonsentsjane 