AI cost so high Google turns cash negative for first time in 22 years
Google has reported negative free cash flow for the first time in more than two decades as AI spending continues to surge. The company says it plans to invest even more in data centres and chips, signalling that its AI push is only just accelerating.
For the first time in more than 20 years, Google has reported negative free cash flow. No, the company is not losing money or facing a financial crunch. Instead, it is spending cash at an unprecedented pace to stay ahead in the AI race, pouring billions into new data centres, AI chips and the computing infrastructure needed to power models like Gemini. While those investments aren't hurting Google's business, they are putting increasing pressure on its cash flow.
Alphabet, Google's parent company, reported negative free cash flow of $5.9 billion in the second quarter. It also raised its 2026 capital expenditure forecast to between $195 billion and $205 billion, up from its earlier estimate of $180 billion to $190 billion. Simply put, Google plans to spend even more on AI infrastructure in the months ahead.
What is free cash flow?
So, if Google isn't losing money, why has its free cash flow turned negative?
To understand that, let's first look at what free cash flow actually means.
Free cash flow is the cash a company has left after paying its operating expenses and investing in long-term assets such as buildings, equipment and infrastructure. In Google's case, that figure slipped into negative territory because the company is spending billions on AI data centres, custom chips, servers and networking equipment needed to train and run its Gemini AI models and meet growing demand for cloud services.
Put simply, Google is still making billions of dollars. It's just spending those billions even faster to build the infrastructure it believes will define the future of AI.
On the company's earnings call, Alphabet Chief Financial Officer Anat Ashkenazi said the pressure on cash flow is expected to continue as Google expands its AI capabilities.
"We expect that free cash flow will remain under pressure driven by our investments in technical infrastructure, which enable us to capitalise on the AI opportunity and continue to drive attractive returns," she said.
Tech giants are pouring billions into AI
Google isn't alone. Across the technology industry, the biggest AI players are spending at an unprecedented pace. Google, Amazon, Microsoft and Meta are collectively expected to spend more than $700 billion on capital expenditure this year, with much of that going towards AI infrastructure.
According to analysts cited by Bloomberg, Google's capital expenditure could rise to at least $262 billion by 2027, while Ashkenazi said spending would increase "significantly" again next year.
And it's not just because companies are spending more. Analysts say the cost of building AI infrastructure is also rising. Memory chips and other critical components have become more expensive, making it costlier to expand data centres than it was a year ago.
When will AI start paying back?
While tech companies are spending hundreds of billions of dollars to build AI infrastructure and stay ahead of the competition, investors are increasingly asking one question: When will these investments start paying off?
The sheer scale of spending has fuelled concerns about how quickly AI will generate meaningful returns. Expectations had already been building that the free cash flow of the four major AI hyperscalers—Google, Amazon, Microsoft and Meta—could turn negative by 2027 if spending continues at the current pace.
Google's business is still growing
Meanwhile, despite the pressure on cash flow, Google's underlying business is going strong. The company reported 24 per cent year-on-year revenue growth, while its cloud business continued to benefit from surging AI demand. Google Cloud revenue jumped 82 per cent to $24.77 billion during the quarter.
For the first time in more than 20 years, Google has reported negative free cash flow. No, the company is not losing money or facing a financial crunch. Instead, it is spending cash at an unprecedented pace to stay ahead in the AI race, pouring billions into new data centres, AI chips and the computing infrastructure needed to power models like Gemini. While those investments aren't hurting Google's business, they are putting increasing pressure on its cash flow.
Alphabet, Google's parent company, reported negative free cash flow of $5.9 billion in the second quarter. It also raised its 2026 capital expenditure forecast to between $195 billion and $205 billion, up from its earlier estimate of $180 billion to $190 billion. Simply put, Google plans to spend even more on AI infrastructure in the months ahead.
What is free cash flow?
So, if Google isn't losing money, why has its free cash flow turned negative?
To understand that, let's first look at what free cash flow actually means.
Free cash flow is the cash a company has left after paying its operating expenses and investing in long-term assets such as buildings, equipment and infrastructure. In Google's case, that figure slipped into negative territory because the company is spending billions on AI data centres, custom chips, servers and networking equipment needed to train and run its Gemini AI models and meet growing demand for cloud services.
Put simply, Google is still making billions of dollars. It's just spending those billions even faster to build the infrastructure it believes will define the future of AI.
On the company's earnings call, Alphabet Chief Financial Officer Anat Ashkenazi said the pressure on cash flow is expected to continue as Google expands its AI capabilities.
"We expect that free cash flow will remain under pressure driven by our investments in technical infrastructure, which enable us to capitalise on the AI opportunity and continue to drive attractive returns," she said.
Tech giants are pouring billions into AI
Google isn't alone. Across the technology industry, the biggest AI players are spending at an unprecedented pace. Google, Amazon, Microsoft and Meta are collectively expected to spend more than $700 billion on capital expenditure this year, with much of that going towards AI infrastructure.
According to analysts cited by Bloomberg, Google's capital expenditure could rise to at least $262 billion by 2027, while Ashkenazi said spending would increase "significantly" again next year.
And it's not just because companies are spending more. Analysts say the cost of building AI infrastructure is also rising. Memory chips and other critical components have become more expensive, making it costlier to expand data centres than it was a year ago.
When will AI start paying back?
While tech companies are spending hundreds of billions of dollars to build AI infrastructure and stay ahead of the competition, investors are increasingly asking one question: When will these investments start paying off?
The sheer scale of spending has fuelled concerns about how quickly AI will generate meaningful returns. Expectations had already been building that the free cash flow of the four major AI hyperscalers—Google, Amazon, Microsoft and Meta—could turn negative by 2027 if spending continues at the current pace.
Google's business is still growing
Meanwhile, despite the pressure on cash flow, Google's underlying business is going strong. The company reported 24 per cent year-on-year revenue growth, while its cloud business continued to benefit from surging AI demand. Google Cloud revenue jumped 82 per cent to $24.77 billion during the quarter.
For the first time in more than 20 years, Google has reported negative free cash flow. No, the company is not losing money or facing a financial crunch. Instead, it is spending cash at an unprecedented pace to stay ahead in the AI race, pouring billions into new data centres, AI chips and the computing infrastructure needed to power models like Gemini. While those investments aren't hurting Google's business, they are putting increasing pressure on its cash flow.
Alphabet, Google's parent company, reported negative free cash flow of $5.9 billion in the second quarter. It also raised its 2026 capital expenditure forecast to between $195 billion and $205 billion, up from its earlier estimate of $180 billion to $190 billion. Simply put, Google plans to spend even more on AI infrastructure in the months ahead.
What is free cash flow?
So, if Google isn't losing money, why has its free cash flow turned negative?
To understand that, let's first look at what free cash flow actually means.
Free cash flow is the cash a company has left after paying its operating expenses and investing in long-term assets such as buildings, equipment and infrastructure. In Google's case, that figure slipped into negative territory because the company is spending billions on AI data centres, custom chips, servers and networking equipment needed to train and run its Gemini AI models and meet growing demand for cloud services.
Put simply, Google is still making billions of dollars. It's just spending those billions even faster to build the infrastructure it believes will define the future of AI.
On the company's earnings call, Alphabet Chief Financial Officer Anat Ashkenazi said the pressure on cash flow is expected to continue as Google expands its AI capabilities.
"We expect that free cash flow will remain under pressure driven by our investments in technical infrastructure, which enable us to capitalise on the AI opportunity and continue to drive attractive returns," she said.
Tech giants are pouring billions into AI
Google isn't alone. Across the technology industry, the biggest AI players are spending at an unprecedented pace. Google, Amazon, Microsoft and Meta are collectively expected to spend more than $700 billion on capital expenditure this year, with much of that going towards AI infrastructure.
According to analysts cited by Bloomberg, Google's capital expenditure could rise to at least $262 billion by 2027, while Ashkenazi said spending would increase "significantly" again next year.
And it's not just because companies are spending more. Analysts say the cost of building AI infrastructure is also rising. Memory chips and other critical components have become more expensive, making it costlier to expand data centres than it was a year ago.
When will AI start paying back?
While tech companies are spending hundreds of billions of dollars to build AI infrastructure and stay ahead of the competition, investors are increasingly asking one question: When will these investments start paying off?
The sheer scale of spending has fuelled concerns about how quickly AI will generate meaningful returns. Expectations had already been building that the free cash flow of the four major AI hyperscalers—Google, Amazon, Microsoft and Meta—could turn negative by 2027 if spending continues at the current pace.
Google's business is still growing
Meanwhile, despite the pressure on cash flow, Google's underlying business is going strong. The company reported 24 per cent year-on-year revenue growth, while its cloud business continued to benefit from surging AI demand. Google Cloud revenue jumped 82 per cent to $24.77 billion during the quarter.
