REDMOND, Wash. (AP) -- Microsoft Corp. said Wednesday that its profit for the October-December quarter grew 10% from the same time last year as it works to capitalize on the huge amounts of money it has spent to advance its artificial intelligence technology.

But while its overall profits and revenue beat Wall Street expectations, it slightly missed projections for its closely-watched cloud computing business, a centerpiece of its AI efforts.

The company reported net income for the quarter of $24.1 billion, or $3.23 per share, beating Wall Street expectations of $3.11 per share. The Redmond, Washington-based software maker posted revenue of $69.6 billion in the quarter, up 12% from the previous year, also beating expectations.

Analysts polled by FactSet Research expected Microsoft to generate revenue of $68.87 billion in the last three months of the year, and currently project revenue of $69.81 billion for the January-March quarter.

Sales from Microsoft's cloud-focused business segment grew 19% from the same time last year to $25.5 billion, which was less than the $25.83 billion forecast by FactSet analysts.

Microsoft's productivity business segment, which includes its Office suite of email and other workplace products, grew 14% to $29.4 billion.

Its personal computing business, led by its Windows division, remained steady at $14.7 billion.

Microsoft shares dropped slightly in after-hours trading Wednesday but were still higher than Monday, when the tech giant was hit by a broader tech stock sale caused by a frenzy over the new ChatGPT competitor developed by Chinese tech startup DeepSeek.

Microsoft is a close partner of ChatGPT maker OpenAI and also sells its own AI chatbot services, branded as Copilot. Part of what drove the Wall Street panic this week was concern that the startup's claims that it was catching up to U.S. tech titans on a fraction of their budget.

Building and operating AI systems is costly, and Microsoft has said it plans to spend $80 billion this year as it expands its global network of energy-hungry computing centers and supplies them with specialized chips to train and run AI models.