Giants in the tech industry are reporting earnings for the first time in 2025 to much anticipation. Investors have been eager to hear how hyperscalers will respond to DeepSeek, which shook markets. Many big tech and semiconductor names sold off extensively after a report that DeepSeek developed its R1 model for under $6 million.
This cast doubt on the need to spend billions to develop AI infrastructure, leading to speculation that hyperscalers could greatly reduce certain investments in AI going forward. However, according to Microsoft (NASDAQ: MSFT) and Meta Platforms (NASDAQ: META), the AI investment train will continue chugging, at least for now.
Below, I’ll break down the important news relating to AI investments revealed by these Magnificent Seven stocks. I’ll also detail commentary from their top management teams explaining what impact they see DeepSeek having. I’ll close by sharing what the developments ultimately mean for these two hyperscalers.
Microsoft: $80 Billion Investment Plan Hasn’t Budged
Microsoft stated at the beginning of Jan. that it is looking to invest $80 billion in fiscal year 2025 to further AI. Specifically, the investment looks to “build out AI-enabled data centers to train AI models and deploy AI and cloud-based applications around the world."
With the company’s fiscal Q2 2025 results coming out on Jan. 30, the fiscal year is already halfway over for Microsoft. So far, the company has spent $43 billion on capital expenditures (CapEx), much of which has likely gone toward AI initiatives. Q1 saw $20 billion in spending, while Q2 had $22.6 billion, an acceleration of 13%. The company still has a lot left to spend on the year. It would be a 60% increase from its $50 billion in CapEx in fiscal 2024.
Microsoft’s Chief Financial Officer indicated that the company could actually surpass its spending target for the full year. She expects CapEx expenditures in Q3 and Q4 to remain at similar levels to Q2. Given that Q2 saw $22.6 billion in capex, the company’s full-year total would come in at $87 billion.
When it comes to DeepSeek, Microsoft appears to be embracing its innovations rather than pushing back. The company has added DeepSeek’s R1 model to its Azure AI Foundry, where customers can use it to build and design apps using over 1,800 models. Microsoft explained that as AI computing prices fall, more AI applications will be written. This means more demand for Microsoft Azure infrastructure to design and run these applications.
Additionally, if AI models require less computing power, they can run locally on devices without relying on data centers. Microsoft is now developing the capabilities for its Copilot+ PCs to run DeepSeek R1 locally. This strengthens Microsoft’s Copilot offering, allowing it to possibly draw higher revenue over time.
Meta: Looking to Learn from DeepSeek's Advances
Just a few days before the DeepSeek shock, Meta said it plans to spend between $60 billion and $65 billion on AI CapEx. It is also looking to grow its AI personnel. The company reiterated this expected pace of investment on its earnings call.
A notable difference between Meta and Microsoft is that Meta’s recent earnings were for the full year 2024. So, we don’t yet have insight into how their stated 2025 spending is shaping up in reality. However, the company said it plans to invest "hundreds of billions" in AI infrastructure over the long term.
In terms of DeepSeek, Chief Executive Officer Mark Zuckerberg explicitly mentioned that the DeepSeek model is a competitor to Meta’s Llama model. He added that there are advancements in the DeepSeek model that Meta will look to put into the Llama model. He mentioned that adopting advancements from competitors isn’t unique to DeepSeek; that’s just how the tech industry works.
He went on to say that it is too early to know what DeepSeek will mean for the company’s trajectory on AI infrastructure and CapEx spending going forward. The company still believes that massive data center investment will pay off. The infrastructure will be necessary to serve billions of users with AI capabilities.
Final Thoughts
Overall, it's largely business as usual when it comes to the AI journeys for these firms. Evolution and pivoting are happening, but earth-shaking changes have not materialized. As was so before DeepSeek, they will need to make more progress in showing the return on investment for their AI projects. The picture of which companies will emerge as the ultimate winners of AI remains blurry, to say the least.
Where Should You Invest $1,000 Right Now?
Before you make your next trade, you'll want to hear this.
MarketBeat keeps track of Wall Street's top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis.
Our team has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on... and none of the big name stocks were on the list.
They believe these five stocks are the five best companies for investors to buy now...
