Microsoft is set to report its second-quarter earnings on Wednesday, with investors closely watching how the company addresses concerns over AI spending and competition from China’s DeepSeek. The emergence of DeepSeek’s low-cost AI model has raised questions about whether the high capital expenditures on AI infrastructure, including Microsoft’s heavy investment in GPUs, remain justified.
AI spending in focus
DeepSeek’s R1 model has shifted the conversation around AI costs, demonstrating that powerful AI systems can be trained at a fraction of the cost previously assumed. This has sparked investor fears that major tech firms like Microsoft, Google, and OpenAI-backed startups may need to rethink their spending strategies.
Despite these concerns, analysts suggest that lower AI infrastructure costs could ultimately benefit Microsoft. “Since Microsoft is focused on Copilots (Gen AI applications) and inferencing, any decrease in the cost of building models or using models is likely positive,” wrote Bernstein analyst Mark Moerdler, according to MarketWatch. Lower costs could improve Microsoft’s AI product margins, potentially driving wider adoption.
Microsoft CEO Satya Nadella has already responded to DeepSeek’s rise, stating that large language models will eventually become a commodity. Analysts expect him to reinforce this view during the earnings call, emphasizing that Microsoft’s AI growth strategy is not solely dependent on training massive models but also on deploying AI-driven services through Azure.
Microsoft’s cloud business and AI investments
Microsoft’s cloud division, particularly Azure, remains central to its AI ambitions. Analysts forecast Azure’s revenue to have grown 32% in the past quarter, with expectations of 33.4% growth in the upcoming quarter. While still expanding, Microsoft’s cloud growth has lagged behind competitors, with Amazon Web Services and Google Cloud posting stronger acceleration in recent months.
One key development is Microsoft’s evolving relationship with OpenAI. Previously, OpenAI exclusively used Microsoft’s Azure cloud for its training and deployment needs. However, Microsoft recently announced that OpenAI will now have to request additional computing capacity rather than having unrestricted access. While this change may improve Microsoft’s capital efficiency, it could also shift some AI workloads away from Azure.
Investor sentiment and the path forward
Microsoft’s stock performance has been relatively weak compared to other tech giants. While Amazon and Google have surged 44% and 26% in the past year, Microsoft has gained just 8%, making it the worst performer among the major tech firms.
However, some analysts see an opportunity for a “mini revenge trade.” Evercore ISI’s Kirk Materne believes Microsoft’s stock could rebound as AI sales increase, Copilot adoption improves, and capital expenditure moderates, leading to stronger cash flow.
Beyond AI, Microsoft is also pushing AI-powered PCs under its Copilot+ PC initiative, aiming to integrate AI functionalities directly into consumer hardware. However, market demand for AI-driven PCs remains uncertain, as consumers continue to prioritize traditional performance improvements over AI features.
With its earnings report, Microsoft will have to convince investors that its AI strategy is sustainable, that its cloud business remains competitive, and that its infrastructure investments will yield long-term returns. DeepSeek’s rise may have raised doubts, but Microsoft’s ability to adapt will be the real test.
Featured image credit: Demure Storyteller/Unsplash
