Detailed Analysis
Alphabet and Amazon's first-quarter 2026 earnings reports revealed a striking structural anomaly in the narrative of AI-driven profit growth: a substantial portion of each company's record-breaking net income derived not from their core products or services, but from paper gains on their equity stakes in Anthropic, the AI safety-focused startup backed by both tech giants. Alphabet disclosed that approximately $28.7 billion of its 81% profit surge to $62.6 billion stemmed from upward revaluations of private company investments, primarily its estimated 14% stake in Anthropic. Amazon was even more explicit, stating directly in its earnings release that first-quarter net income included $16.8 billion in pretax gains from its Anthropic investment — a figure representing more than half of the company's total pretax income for the quarter. These disclosures arrived in the same reporting cycle in which the four largest U.S. tech companies collectively announced $130.65 billion in capital expenditures for a single quarter, a figure the article notes exceeds three times the inflation-adjusted cost of the Manhattan Project.
The nature of these gains demands close scrutiny. Both figures represent non-operating, mark-to-market valuation adjustments — unrealized paper profits that reflect how each company has re-priced the present value of its Anthropic equity as the startup's implied valuation has risen sharply. Amazon's position is particularly striking: an $8 billion confirmed investment now carries an estimated value of $60.6 billion, a roughly sevenfold return, generating recognized accounting gains of over $22 billion between Q3 and Q4 2025 alone, with an additional projected $15 billion in Q1 2026 from note conversions. Google's $3 billion investment for its 14% stake, plus a recently announced commitment of up to $40 billion more, has similarly appreciated as Anthropic's valuation has climbed. These are not revenue-generating transactions; they are balance sheet events that flow through to net income under standard accounting rules, making headline profit figures appear more robust than operational performance alone would suggest.
The deeper context is that both cloud giants have structured their Anthropic relationships to generate compounding strategic advantages beyond equity appreciation. Anthropic has committed to purchasing more than $100 billion in compute from Amazon Web Services over a decade, and has similarly concentrated significant cloud infrastructure spend with Google Cloud. Industry analysts estimate that cloud providers capture between 16% and 32% of Anthropic's revenue through high-margin infrastructure fees, effectively monetizing the startup's explosive growth — projected at $5 billion annualized in 2025 and potentially $26 billion by 2026 — without bearing the risk of frontier model development. This means the Anthropic relationship functions as a multi-layered profit engine: equity appreciation, locked-in cloud consumption, and accelerated enterprise AI adoption driven by Anthropic's model deployments all benefit the investing companies simultaneously.
The broader implications of this earnings structure touch on a question that has become central to AI investment analysis: how much of the current profit boom reflects durable operational value creation versus financial engineering and speculative asset appreciation. The $700 billion in planned 2026 capex across the largest tech firms — equivalent to the entire U.S. Medicare budget — is being justified in part by earnings figures that are themselves substantially derived from the rising valuations of private AI companies, creating a potential circularity in the investment thesis. If Anthropic's private valuation were to stagnate or decline, those non-operating gains would reverse into losses, potentially revealing a more modest underlying profit trajectory from core cloud and advertising businesses. This dynamic echoes historical moments in technology investment cycles where infrastructure spending and asset appreciation became mutually reinforcing until an external shock interrupted the feedback loop.
The Anthropic stake disclosures also illuminate the unusual competitive structure now defining the frontier AI industry. Rather than building rival models entirely in-house, the world's largest technology companies have effectively placed enormous financial bets on an external AI developer while simultaneously using that relationship to channel the startup's compute demand back into their own cloud infrastructure. This arrangement concentrates risk at Anthropic while distributing the upside broadly across its investors and cloud partners — a structure that resembles a strategic hedge more than a conventional venture investment. As Anthropic's valuation and revenue continue to scale, the dependency between the company's growth and its backers' reported financial health will only deepen, making Anthropic's trajectory an increasingly material variable in the earnings outlooks of two of the world's most valuable public companies.
Read original article →