Microsoft & Meta said that their big spending in AI is justified

After Chinese startup DeepSeek launched its cost-effective AI that was built with just $6 million and older Nvidia H100 GPUs, the CEOs of Microsoft and Meta defended their big spending in AI investment.

DeepSeek’s model claimed to rival Western AI at a lower cost, raising concerns about America’s lead in the sector. Despite this, both CEOs said that there’s a need for massive infrastructure spending to meet growing corporate demand as they spoke during recent quarterly earning reports.

Microsoft has committed $80 billion to AI, while Meta plans $65 billion, far surpassing DeepSeek’s $6 million investment. Investors are now impatient with the slow return, as both companies face competition from cheaper Chinese models and unclear monetization paths.

The Redmond tech giant has recently been integrating NPU-optimized DeepSeek R1 models into its Copilot+ PCs, starting with Qualcomm’s Snapdragon X chips and later expanding to Intel Core Ultra 200V.

These models will run locally on devices using the Windows Copilot Runtime and AI Toolkit.

Claiming to have developed its powerful AI less resources, DeepSeek’s rise has caused a big panic and market disruption, with the US stock market losing over a trillion dollars. Both Microsoft & Meta were among the tech firms that suffered in the stock market, albeit in recovery.

Analyst Charles Archer suggests three possible reasons for DeepSeek’s success: reverse-engineering open-source AI, Chinese government-backed funding, or third, innovative scaling with older hardware, which could be huge, if true.

Despite concerns over DeepSeek’s rise, however, Microsoft posted strong results for the latest quarter, with a 14% growth in business revenue and a 19% increase in its cloud segment. The company’s total revenue also reached $69.6 billion, fueled by Azure’s 31% revenue growth.