Google's AI bet is paying off: Alphabet posts $94.7 billion Q1 revenue, AI demand outpaces supply

Skye Jacobs

Posts: 2,010   +58
Staff
Bottom line: Alphabet's latest results show how central artificial intelligence has become to its growth – and how aggressively the company is increasing spending to support the technology. The Google parent reported $94.7 billion in first-quarter revenue, excluding partner payouts, beating analysts' expectations of $91.6 billion. Earnings came in at $5.11 per share, well above the $2.62 estimate. Investors responded quickly, pushing the stock up more than 6% in after-hours trading.

Much of Alphabet's momentum is tied to AI, particularly inside Google Cloud. The division brought in $20 billion in revenue, ahead of the $18.4 billion analysts had projected, and showed what the company called a "meaningful acceleration in growth." Just as notable, its backlog – future revenue under contract – nearly doubled from the previous quarter to more than $460 billion, reflecting growing demand for AI infrastructure and services.

That demand is strong enough that Google can't fully keep up. "We are compute constrained in the near term, and as an example, our cloud revenue would have been higher if we were able to meet the demand," CEO Sundar Pichai said. "We are working through that moment and investing."

The scale of that investment is significant. Alphabet now expects to spend up to $190 billion on capital expenditures this year, slightly above its prior estimate and roughly double its 2025 spending. Chief Financial Officer Anat Ashkenazi said spending will climb further in 2027. "These strong results reinforce our conviction to invest the capital required to continue to capture the AI opportunity," she said.

Credit: App Economy Insights

A significant portion of that spending is going toward hardware. Google's tensor processing units, or TPUs, have long powered its internal AI workloads, but the company is now preparing to offer them directly to select customers for use in their own data centers. The move puts TPUs in more direct competition with Nvidia's chips as companies seek scarce high-performance compute.

On the software side, adoption is rising alongside infrastructure. Google said use of its Gemini models continues to expand across both consumer and enterprise products. The Gemini chatbot reached 750 million users by the end of 2025, while paid monthly active users for Gemini Enterprise grew 40% from the previous quarter.

At the same time, Google is continuing to rework its core search business around AI. The company now integrates AI-generated answers directly into many search results, and Pichai said query volume has reached an all-time high since those features were introduced. He also noted that Google has reduced the cost of serving AI-powered responses, addressing earlier concerns that generative AI could make search significantly more expensive to operate.

The results help address investor concerns that chatbots from companies like OpenAI and Anthropic could erode Google's search dominance. "AI has enhanced search, not killed it," Andrew Rocco, a strategist at Zacks Investment Research, wrote. "Google has masterfully integrated AI into its search offering."

Even so, the competitive landscape is becoming more complicated. Alphabet is both a backer of and a rival to some of the companies competing in the same AI markets. It has invested heavily in Anthropic, with plans to put up to $40 billion into the startup, while also competing against it in areas like AI coding tools. Anthropic's Claude Code, in particular, has emerged as a fast-growing product that has drawn attention inside Google.

For now, Alphabet's results indicate that its strategy – spanning chips, cloud, and AI applications – is gaining traction. But the company's own comments about limited compute capacity point to a constraint that extends beyond Google. Across the industry, scaling AI is becoming as much an infrastructure challenge as a software one.

Permalink to story:

 
Imagine if Google had embraced generative AI when their own researchers first invented the transformers architecture (the "Attention is all you need" paper). They would be light years ahead at this point and probably patent-litigating against anyone else who tried to do it.
 
Wait.. what ?????
Record revenues, profits crush expectations, AI demand outpaces supply ... how is that possible?
I mean, according to the experts we're in a bubble. The studies conducted by the same experts tell us the attempts to use AI are failing left and right ..

Whom to trust, facts or experts?
 
In other news: If you steal other people's stuff without paying them, then you can increase your profits while driving the original producers out of business. At least in the short or medium term.
 
Record revenues, profits crush expectations, AI demand outpaces supply ... how is that possible? I mean, according to the experts we're in a bubble. The studies conducted by the same experts tell us the attempts to use AI are failing left and right ..
There is no contradiction between these statements. Actually, in every bubble demand usually outpaces supply - until it pops, at which time there will be an oversupply of the previously overvalued product, and that's the real problem causing the actual crash.

That being said "AI demand outpacing supply" is nowhere actually found in Google's financial report to begin with. It's just a false interpretation of the actual numbers and facts in the report. Which only show that Google Cloud revenue (as that of a few other divisions') has increased.

But it's not even evident that that has anything to do with AI and the demand for it. And even if it would have to do with that, that could still be for ex. because OpenAI or other AI companies are running their services fully or partially on Google Cloud, which in turn generates a tons of revenue for Google, despite the respective AI companies themselves running only losses.
 
Wait.. what ?????
Record revenues, profits crush expectations, AI demand outpaces supply ... how is that possible?
I mean, according to the experts we're in a bubble. The studies conducted by the same experts tell us the attempts to use AI are failing left and right ..

Whom to trust, facts or experts?
They are making record money selling server space to other companies that want to run AI for themselves or to sell to others. -

They are not making money selling their AI.
 
Imagine if Google had embraced generative AI when their own researchers first invented the transformers architecture
Imagine if AT&T had embraced its development of the transistor back in 1950, instead of selling the rights to Sony for a few thousand dollars.

BTW, transformer models are just as 'transformative' for predictive AI as they are generative, even though LLMs get all the media attention.
 
They are making record money selling server space to other companies that want to run AI for themselves or to sell to others. -

They are not making money selling their AI.
Nailed it! Google's making money selling the shovels, not the gold.

And that "we are compute constrained" or "we'd have made more money if we had more chips" line is the trending corporate flex of the year that all shareholders want to hear.
 
Back