Microsoft Earned $3.2 Billion on Its Anthropic Stake, but OpenAI Delivered a Mixed Result

On July 29, 2026, in its report for the fourth quarter of fiscal 2026 (which ended June 30), Microsoft disclosed the results of its investments in two competing AI labs: its stake in Anthropic generated $3.2 billion in accounting profit, while its investment in OpenAI delivered a mixed result.

Where the $3.2 billion came from

Microsoft earned the $3.2 billion in profit from its stake in Anthropic, the company disclosed in its quarterly report for the period that ended June 30, 2026. The figure is recorded under the equity method: the corporation reflects in its own financials a proportional share of the results of companies in which it holds a significant but non-controlling stake. The rise in Anthropic's value and results turned into a concrete profit line for Microsoft rather than remaining a paper valuation for the future.

Notably, Microsoft did not put the figure in the headline of the report. It tucked it inside quarterly results that were themselves strong. The disclosure shows how bets on AI startups are gradually moving out of the "strategic spending" category and into real financial results.

• Reporting period: Microsoft's fiscal Q4 2026, which ended June 30, 2026
• Profit from the Anthropic stake: $3.2 billion
• Accounting method: equity method
• Investment in OpenAI: mixed result
• Anthropic and OpenAI: the two largest competing AI labs

Why Microsoft is betting on both rivals

Microsoft has ended up as a co-owner of two direct rivals: Anthropic (developer of the Claude model family) and OpenAI (creator of ChatGPT). For OpenAI, the corporation is a longtime and its largest outside partner, while it joined Anthropic's capital later. The dual bet hedges the Big Tech giant against the risk that only one lab takes the lead in frontier models: whichever product wins the race, Microsoft comes out ahead as a shareholder on both sides.

The financial stakes also complement a product dependency: Microsoft plugs the models of both labs into its services, from the Azure cloud to the Copilot lineup. That makes the $3.2 billion from Anthropic more than a one-off revaluation. It is part of a broader bet in which the corporation is both an investor and the largest distribution channel for these models.

Why OpenAI delivered a mixed result

According to TechCrunch's wording, the OpenAI investment turned out to be a "mixed result", unlike the clear gain from Anthropic. The equity method works in both directions: when a portfolio company posts losses, the investor also reflects its share of those losses in its own reporting. OpenAI remains deeply unprofitable and is increasing its compute spending faster than its revenue, so its contribution to Microsoft's balance sheet for the quarter that ended June 30, 2026 looks different from the profit from Anthropic.

"Microsoft earned $3.2 billion on its investment in

Anthropic, but with OpenAI the result was mixed," reads TechCrunch's headline on the company's quarterly results.

What this means

The report shows more clearly than ever that Big Tech's AI investments are no longer a pure expense line and are starting to show up in profit, though so far unevenly. Anthropic is already bringing Microsoft billions on paper, while OpenAI is burning cash to scale. For investors, this is the first measurable gauge of what Microsoft's AI bets are really worth, and a signal that stakes in the leaders of generative AI are starting to work as full-fledged financial assets.

As both labs raise new rounds at ever higher valuations, lines like this will appear more often in Big Tech earnings reports, and the market will start judging the AI race not only by model releases but also by the numbers in quarterly filings.