Microsoft’s Q2 AI Business Soars but Cloud, Guidance Disappoint

Microsoft reported robust growth in artificial intelligence (AI) demand for the second quarter, with gains seen throughout its Copilot and Azure operations and propelling revenue and earnings past Wall Street expectations.

But non-AI business for its cloud division, Azure, disappointed Wall Street because it came in at the low end of expectations. Microsoft also expects this segment to be impacted throughout the second half of its fiscal year. Shares of Microsoft were down 6% in late morning trading Thursday (Jan. 30).

CEO Satya Nadella said Microsoft’s AI business has now exceeded an annual revenue run rate of $13 billion, up 175% year over year. The metric extrapolates what yearly revenue would be at the current growth pace.

“Enterprises are beginning to move from proofs-of-concept to enterprise-wide deployments to unlock the full ROI of AI,” Nadella said.

In the past two years, as ChatGPT brought a surge of excitement about generative AI, companies have been struggling to deploy the technology in their business, often not moving past the proof-of-concept stage.

But momentum is building. Nadella sees more businesses deploying AI as costs come down: “As AI becomes more efficient and accessible, we will see exponentially more demand.”

Nadella said DeepSeek’s reasoning model, R1, will be added to the Microsoft Azure model lineup and made accessible to customers. Soon, DeepSeek will be able to run at the edge, stored and processing locally on Windows-based AI PCs, without having to access the cloud or the internet.

DeepSeek roiled shares of Nvidia and other chipmakers on Monday (Jan. 27) after disclosing that it spent $5.6 million and used 2,000 older Nvidia H800 chips to develop one of its foundation models. In contrast, GPT-4 reportedly had cost over $100 million to train.

Silicon Valley took note of DeepSeek’s engineering innovations that let it cut training costs significantly.

“DeepSeek has had some real innovations,” Nadella said.

However, he pointed out that the AI industry has become more efficient and lowered prices as well, helped by two laws of technology.

“There’s Moore’s Law that’s working in hyperdrive. Then on top of that, there’s the AI scaling laws,” Nadella said.

AI scaling laws describe how AI models improve performance with more data and computation. Moore’s Law, although slowing down, states that the number of transistors on a microchip doubles about every two years.

These two combined lead to more powerful chips and more efficient AI models, resulting in lower costs and rising performance, Nadella said.

When the cost to use deployed AI models (inference) come down, “that means people consume more, and there will be more apps written,” he said. “That type of optimization means AI will be much more ubiquitous.”

“Therefore, for a hyperscaler like us, a PC platform provider like us, this is all good news as far as I’m concerned,” Nadella said.

When it comes to AI agents, bots that not only provide information but also complete task autonomously for the user, Microsoft sees growth, too.

The company said more than 160,000 organizations have already used Microsoft’s Copilot Studio to create more than 400,000 custom agents in the last three months. That is more than double from the prior quarter.

While the company remains bullish about AI, its non-AI business for Azure — those from clients coming in through third parties — will see headwinds into the rest of Microsoft’s fiscal year. Microsoft noted it had “execution challenges” as it balances its AI and non-AI business.

However, Nadella added that “you would rather win the new than just protect the past” and that’s a strategy “we definitely will lean into always.”

For the second quarter, Microsoft reported net income of $24.1 million, or $3.23 per share, beating Wall Street consensus estimates of $3.11 per share and up 10% year over year. In the same quarter a year ago, Microsoft earned $21.9 million, or $2.93 per share.

The software giant posted revenue of $69.63 billion, up 12% from the same quarter a year ago. Analysts were expecting $68.7 billion.

But shares were under pressure in late morning trading on Thursday (Jan. 30).

BofA Global Research analyst Brad Sills wrote in a note that “rare execution challenges impact Azure.” This led to second-quarter revenue growth that was at the “low end of guidance” due to issues in the broad partner channel.

“The issues appear fixable, though will impact revenue in the coming quarters,” the analyst said.

Sills kept his “buy” rating on the stock with a price target of $510. Microsoft is currently trading at $415.48.

“We continue to believe that Microsoft remains uniquely positioned to monetize the vast new AI opportunity across applications and infrastructure at scale,” Sills said.

The post Microsoft’s Q2 AI Business Soars but Cloud, Guidance Disappoint appeared first on PYMNTS.com.