Microsoft’s latest earnings slipped in a telling subplot: one AI bet is suddenly paying off handsomely, while another looks more complicated. The result is a sharper picture of how Big Tech is trying to profit from the generative AI boom without backing only one horse.

In the quarter ending June 30, Microsoft booked a $3.2 billion gain on its Anthropic investment, a boost that added 33 cents to diluted earnings per share.¹ That is striking on its own, but the contrast is what makes it newsy: in the same period, Microsoft marked down its OpenAI stake by roughly $600 million, trimming EPS by about 7 cents.¹ TechCrunch summed it up neatly: Microsoft’s AI portfolio was lucrative, but “OpenAI was a mixed bag.”¹

That does not mean Microsoft is souring on OpenAI. On a full-year basis, its OpenAI investment still produced a $5 billion gain, suggesting the quarterly markdown was more wobble than rupture.¹ Still, the optics matter. Microsoft disclosed nearly as much gain from Anthropic in one quarter as it got from OpenAI over the full fiscal year, a reminder that the company’s hedge across rival AI labs is starting to look deliberate rather than incidental.¹

And Microsoft is not alone. Amazon also told investors that Anthropic is turning into a financial jackpot, reporting $53.4 billion in non-operating pre-tax other income “primarily from our investment in Anthropic.”² Business Insider framed it as a “huge windfall,” tying that surge to Anthropic’s swelling valuation and IPO preparations.²

The common thread is simple: Anthropic is no longer just an AI partner. For Microsoft and Amazon, it is becoming a balance-sheet event. The difference is in what that says about strategy. For Amazon, it looks like a massive upside bet paying off. For Microsoft, it underscores something more nuanced — the company is making money from AI’s leading contenders even as the race between them stays unsettled.