In the wake of a breakthrough by Chinese upstart DeepSeek in affordable AI computing, executives at Microsoft and Meta have defended their significant investments in AI infrastructure, despite growing concerns over mounting costs and investor impatience.
DeepSeek, a rising player in the AI space, has stunned the tech industry with claims that its models can compete with, and even surpass, Western counterparts, all at a fraction of the cost. This rapid progress has sparked doubts about the U.S.’s dominant position in AI, with some observers questioning whether the enormous capital investments made by U.S. tech giants can continue to guarantee a competitive edge.
Meta CEO Mark Zuckerberg emphasised the importance of investing heavily in capital expenditure, insisting that such spending would ultimately provide a long-term strategic advantage. “Investing ‘very heavily’ in capital expenditure and infrastructure is going to be a strategic advantage over time,” Zuckerberg remarked during a post-earnings call.
Similarly, Satya Nadella, the CEO of Microsoft, underscored the necessity of such investments to meet the increasing demands driven by AI. “As AI becomes more efficient and accessible, we will see exponentially more demand,” he said during a call with analysts.
Microsoft has committed $80 billion to AI in its current fiscal year, while Meta has pledged up to $65 billion. These sums dwarf the $6 million DeepSeek claimed to have spent developing its AI model, though experts argue the discrepancy largely stems from the vast amount spent on computing power rather than the overall development costs.
Meanwhile, Microsoft Corp and OpenAI are also currently investigating whether data generated by OpenAI’s technology was accessed without authorization by a group associated with the Chinese startup DeepSeek, according to sources familiar with the matter, as reported by Bloomberg.
As per the agency report, citing sources, Microsoft security researchers found that individuals, allegedly connected to DeepSeek, extracted a significant volume of data through the OpenAI application programming interface (API).
The report also said that software developers could pay a license fee to use the API to integrate OpenAI’s proprietary artificial intelligence models into their own applications.
