Alphabet, the parent company of Google, has announced plans to invest around $75 billion in capital expenditure in 2025, which will go to growing the company’s fleet of AI-ready data centres.
The race to lead the AI space is definitely heating up, with Microsoft having recently committed to spend $80 billion on the development of new AI data centres, while last month OpenAI announced a partnership with SoftBank, Oracle and MGX which will spend $500 billion on building new AI infrastructure in the United States.
Alphabet understands that in order to continue competing in the space, it will have to follow those companies in spending big. During its earnings call this week, Alphabet CEO Sundar Pichai noted, “We are confident about the opportunities ahead, and to accelerate our progress, we expect to invest approximately $75 billion in capital expenditures in 2025,”
Exactly where Alphabet plans to invest the $75 billion is not exactly clear, but the company has confirmed that “technical infrastructure, primarily for servers, followed by data centres and networking,” would account for a significant portion of the spend, with an estimated $16-18 billion of set to be allocated in the first quarter of 2025.
It wasn’t that long ago that Alphabet announced investment in a $1 billion data centre in Hertfordshire, UK, so it’s possible that the lion’s share of investment for Alphabet’s capital expenditure will be spent in the US – just like Microsoft’s $80 billion commitment. However, with the UK Government looking to attract investment into AI infrastructure in the UK by streamlining planning, Alphabet might be tempted to invest further in the market.
Industry has begun questioning big AI expenditure
Whether all this investment is needed remains to be seen, however. The launch of DeepSeek’s latest AI models, which were developed by Chinese developers, have called into question the cost of developing AI. While American AI companies have committed to spending billions, DeepSeek’s R1 model reportedly cost just $6 million. There are, however, doubts on that figure, with Australia’s Macquarie recently stating that the more likely cost of developing R1 actually stood at $2.6 billion.
There are also growing concerns in Wall Street that spending on AI is too high. In fact, after Alphabet announced its $75 billion capital expenditure commitment, its share price began to slide, dropping around 8%. Of course, that was also due to the company missing its revenue estimates.
Google drops pledge not to use AI for weapons
In a separate development, Google has quietly removed a public commitment to refrain from developing AI for weapons or surveillance from its website, as first spotted by Bloomberg. The updated AI principles no longer list ‘applications we will not pursue’, a section present until recently. Instead, a newly released company blog post states: “we believe that companies, governments, and organisations sharing these values should work together to create AI that protects people, promotes global growth, and supports national security.”
Google’s updated principles also promise to “mitigate unintended or harmful outcomes and avoid unfair bias,” and reference “widely accepted principles of international law and human rights.”
It’s not all too surprising to see Alphabet shift its policy, especially with many of its competitors seeing huge opportunities for growth in the defence sector. OpenAI has recently announced a slew of partnerships with defence contractors, that deal came after it dropped its own pledge to not use AI for military use. Meanwhile, Microsoft’s AI tech was supposedly on hand to help out the Israeli military during its recent incursion into Gaza.
