It's been a tough year for software stocks and hyperscalers. Software investors now fear potential disruption from artificial intelligence model companies, while hyperscalers have sold off after forecasting massive spending plans to build out AI infrastructure.
Count Microsoft (MSFT 0.73%) as both. While some other prominent software stocks and hyperscalers have sold off by more, Microsoft's stock has declined 30% from its all-time highs set back in July. That's significant bear-market action, in-line with the nauseating sell-off Microsoft experienced in 2022, and all the more surprising given Microsoft's perception as one of the "safest" stocks in the market.
Is there anything encouraging or reassuring for Microsoft shareholders on the horizon? One insider thinks so -- and just made a $2 million bet on better times ahead.
NASDAQ: MSFT
Key Data Points
Director John Stanton buys the dip
On Feb. 18, Microsoft director John W. Stanton bought about $2 million worth of stock, purchasing an even 5,000 shares at an average price of $397 per share. The open market purchase increased Stanton's stock holdings by 6.1%.
That might not sound like much, but $2 million is certainly not nothing. And it's especially interesting that Stanton would take the trouble to buy more stock, as he already receives about $250,000 in stock awards as part of his annual compensation as a Microsoft director.
Stanton has been a Microsoft director since 2014, is currently a partner at a major private equity fund, and was a telecom industry executive before then. Clearly, he's pretty familiar with the business and apparently believes Microsoft will get through this period unscathed.
So, what might Stanton see in Microsoft's shares?
Microsoft is somewhat hedged against AI disruption
One saving grace for Microsoft is its roughly 27% stake in OpenAI. Meanwhile, in November, Microsoft announced it was investing "up to" $5 billion in OpenAI rival Anthropic. So, assuming Microsoft is forced to pay a tax to these two leading large language models (LLMs) for incorporating them into its Office or Dynamics software, Microsoft is somewhat "hedged" through its partial ownership of these two companies.
Furthermore, as part of these investment deals, both OpenAI and Anthropic have committed to a massive amount of computing on Microsoft's Azure cloud. In its deal announcement, Anthropic committed to at least $30 billion of compute on Azure over the next few years. And OpenAI has committed a whopping $280 billion to Azure over the near and medium term.
Therefore, Microsoft is hedged in a few ways. If these LLMs develop competing software to Microsoft's, Microsoft will still retain a share of their profits, while also being a cloud vendor to the winners. A more likely case is that, given Microsoft's massive installed base of Microsoft Office 365 customers, these LLMs will probably partner with Microsoft, licensing their software and integrating it with Microsoft, rather than building a competitor.
However, Microsoft is also pursuing a third option, which is likely to preserve its software franchise margins to an even greater extent.
Image source: Getty Images.
Don't forget about Microsoft's own AI ambitions
Unlike virtually all other software vendors, Microsoft is also building its own AI models, specifically to serve its software franchise. In August 2025, Microsoft unveiled its MAI-1 mixture-of-experts model, which actually comes in several model variants for different use cases. In the press release, Microsoft said: "[D]ifferent user intents and use cases will unlock immense value. There will be a lot more to come from this team on both fronts in the near future."
Then in January, Microsoft unveiled its Maia 200 AI inference chip, which it claimed outperformed its other cloud vendors' in-house-designed chips in terms of both TFLOPS and HBM capacity.
So Microsoft is pursuing vertical integration in AI as well, peobably with the goal of serving its software with its own models. That would undoubtedly be the most profitable way to go about things, as it would cut out the margin Microsoft would have to pay OpenAI or Anthropic.
Microsoft has much less risk than other software names
Given that it has three paths to success in the AI era, through its ownership of OpenAI and Anthropic, its position as a cloud vendor to these LLMs, and its pursuit of its own custom models built at least partially on its own hardware, Microsoft has much less risk than other software names, which may see their margins compress in the AI era.
Given that the stock has fallen to nearly its lowest P/E ratio in a decade, no wonder Stanton may see opportunity.
The only real existential risk is if a certain LLM that's not OpenAI or Anthropic, perhaps Google Gemini, achieves some sort of breakthrough that enables it to dominate the market in a monopolistic fashion. But as it stands today, it seems unlikely this will be a winner-take-all market anytime soon.
