The CEO of Google, Sundar Pichai, said his company will invest roughly $75B in capital projects in 2025, most of which will likely go into AI initiatives.
In 2025, Sundar Pichai, the CEO of Google, anticipates that his business will make approximately $75 billion in capital expenditures to expand its artificial intelligence offerings.
During Alphabet’s fourth-quarter 2024 earnings announcement, Pichai stated, “We anticipate investing roughly $75 billion in capital expenditures in 2025.” Compared to the company’s $32.3 billion in 2023 capital expenditures, the amount represents a 43% increase.
He said the investment will “accelerate progress” in AI innovation and keep the company’s core operations strong.
The money spent to buy long-term fixed or physical assets for corporate operations is known as capital expenditures, or “capex.”
Although the precise amount of the investment allocated to AI is unknown, most of it is anticipated to go toward growing Google’s AI infrastructure. Like Meta, which announced it would invest $65 billion to enhance its AI infrastructure, other large tech companies have also upped their investment in AI-related projects.
With total income growing 12% yearly at $96.5 billion, artificial intelligence has recently emerged as one of Google’s most lucrative revenue streams.
In the same period, Google Cloud revenue increased by 10% to $12 billion, which the company claimed was fueled by steady growth in “core Google Cloud Platform products, AI Infrastructure, and Generative AI Solutions.”
However, according to Yahoo Finance, Alphabet’s share price dropped more than 7% in after-hours trading, and its total revenue missed the $96.7 billion combined expert estimates.

Sundar Pichai minimized the dangers of emerging rivals, such as the China-based AI model DeepSeek, which rocked the market in late January, in an investor call on February 4.
Even compared to DeepSeek’s v3 and R1 models, Pichai informed callers that Google’s Gemini 2.0 Flash models are “some of the most efficient models out there.”
The revelation that DeepSeek’s creators have created a competitive alternative to American AI companies like OpenAI at a far lower cost startled US markets on January 27.
The company used less sophisticated gear from semiconductor giant Nvidia to create its AI model on a shoestring budget of less than $6 million.