Why Warren Buffett is betting big on Google as his top AI stock

Warren Buffett has never been known as a tech-chasing investor. For decades he avoided high-flying technology stocks, preferring simple, cash-generating businesses he could understand. That’s why his latest move into artificial intelligence has caught so much attention.

Through Berkshire Hathaway, Buffett has reportedly built a multibillion-dollar position in Alphabet, Google’s parent company. For an investor famous for caution, a roughly $12 billion bet on an AI leader is a strong signal that Google may be one of the most compelling AI stocks in the market today.

The AI stock Buffett is buying: Alphabet (Google)

The centerpiece of this new AI push is Alphabet. While Berkshire’s recent filings also show positions in companies like The New York Times and Delta Air Lines, the standout AI play is clearly Google.

Buffett’s buying reportedly began around March–May of this year, when Alphabet’s share price was significantly lower. Since then, the stock has climbed roughly 30–40%, far outpacing the average annual return of the S&P 500. For someone who has been sitting on hundreds of billions in cash and warning that markets are overvalued, deploying $12 billion into Google is a strong endorsement of its long-term AI potential.

Why Buffett still likes classic business traits

Even when he invests in tech, Buffett’s core principles don’t change. He still looks for:

• Consistent revenue and profit growth
• Strong free cash flow
• Durable competitive advantages (a “moat”)
• A business model he can understand

Alphabet checks all of these boxes. Over the past several years, Google’s revenue has climbed from around $250 billion to roughly $400 billion, with net income growing alongside it. Profit margins have remained high, often near 30%, which is exceptional for a company of its size.

Free cash flow has also been strong, with a temporary dip largely explained by heavy AI investments—data centers, chips, and model training—rather than a weakening core business. For a value-focused investor, spending cash to strengthen a long-term competitive moat can be a feature, not a bug.

Google’s business model: scale without matching costs

One of the main reasons Google is so attractive as an AI stock is its business model. Once the infrastructure is built—data centers, software platforms, and core services—the cost of serving each additional user is relatively low.

Google offers many of its products for free: Search, Gmail, Maps, YouTube, Docs, and more. In return, it collects data and attention, which it monetizes primarily through advertising and, increasingly, cloud and enterprise services.

Because fixed costs are so high and incremental costs are so low, adding more users doesn’t significantly increase expenses. But it does increase ad impressions, data for training AI models, and opportunities to sell premium services. That combination of scale, data, and monetization is exactly the kind of economic engine Buffett likes.

Dominating the internet user base

Google’s reach is staggering. Out of roughly 5.6 billion internet users worldwide, an estimated 5.1 billion use Google services. That’s close to 90% of the online population.

This dominance matters for AI because every user interaction generates data—search queries, clicks, navigation patterns, video views, and more. That data is fuel for training and improving AI models. The more people use Google, the better its models can become, which in turn makes its products more useful, creating a powerful flywheel effect.

In other words, Google hasn’t just won the search game; it has positioned itself to be a default gateway for AI experiences for billions of people.

AI is more than chatbots: Google’s end-to-end stack

When most people think of AI, they think of chatbots like ChatGPT or Gemini. But behind these tools is a complex supply chain: chips, data centers, models, software platforms, and user-facing products. The more of this stack a company controls, the more power and profit potential it has.

Designing its own AI chips

AI requires extremely advanced chips. Google designs its own AI accelerators, known as TPUs (Tensor Processing Units). While it doesn’t manufacture them itself—that’s handled by foundries like TSMC—designing its own chips lets Google optimize hardware specifically for its models and workloads.

Very few companies in the world design their own high-performance chips at this level. That puts Google in a small, elite group alongside players like Apple and Nvidia.

Owning the data center infrastructure

Once chips are designed and manufactured, they need to be deployed at scale. That’s where data centers come in. Google has spent years and tens of billions of dollars building some of the most advanced data center infrastructure on the planet.

These facilities are where Google trains and runs its AI models. By owning and operating its own global network of data centers, Google reduces its reliance on third parties and can tightly integrate hardware, software, and networking for maximum efficiency.

Building frontier AI models: Gemini

On top of this infrastructure sits Google’s family of AI models, branded as Gemini. These large language models and multimodal systems power chat experiences, code generation, content creation, and more.

Gemini is being positioned as a direct competitor to other leading AI models and platforms. If you’re interested in how these kinds of models are changing work and productivity, you may also want to check out our guide on using Claude Co-work more effectively, which shows how advanced AI assistants are reshaping knowledge work.

Integrating AI into every Google product

One of Google’s biggest advantages is the sheer number of products where it can embed AI. Gemini isn’t just a standalone chatbot; it’s being woven into the entire Google ecosystem.

Examples include:

• Search: AI overviews and summaries that answer questions directly on the results page
• Gmail: Smart drafting, summarization, and organization powered by Gemini
• Docs, Sheets, and Slides: AI-assisted writing, analysis, and content creation
• Android and devices: On-device AI (like Gemini Nano) for smarter, more private experiences

As AI becomes more capable, Google can roll new features out to billions of existing users almost instantly. That makes it much easier to monetize AI than for a startup that has to acquire users from scratch.

Powering other companies’ AI transformation

AI won’t just live inside consumer apps. Banks, hospitals, factories, law firms, and countless other organizations are racing to integrate AI into their operations. Most of them don’t have the expertise or infrastructure to build their own models from the ground up.

This is where Google Cloud and Gemini come in. Companies can use Google’s models and infrastructure as a foundation, customizing them for their own needs. Instead of building AI from scratch, they can effectively “rent” Google’s AI capabilities—paying for computing, storage, and model access.

As more industries automate workflows, decision-making, and customer service with AI, platforms like Google Cloud and Gemini stand to capture a growing share of that spending. This aligns with broader AI mega trends and the positioning of top AI stocks discussed in our article on 3 AI mega trends and 10 stocks positioned to win.

How AI could supercharge Google’s financials

If Google successfully executes on its AI strategy, the financial impact could be enormous. There are three main levers:

1. Higher revenue: New AI-powered products and services, more valuable ads, and expanded cloud offerings can all drive top-line growth.
2. Better margins: Automation and AI-assisted operations can reduce internal costs, improving profit margins over time.
3. Stronger free cash flow: As AI investments mature and scale, the combination of higher revenue and efficient operations can translate into significantly higher free cash flow.

Some analysts and investors believe Google could eventually be one of the first companies to reach a $10 trillion valuation if AI dramatically expands its role in the global economy. While that’s speculative, the underlying logic is that controlling key parts of the AI stack and distribution to billions of users could justify a much larger business over time.

What this means for AI investors

Buffett’s move into Alphabet doesn’t mean every investor should blindly follow, and it doesn’t guarantee future returns. But it does highlight a few important ideas for anyone looking at AI stocks:

• AI is not just about chips or chatbots; it’s about full-stack control—from hardware to models to distribution.
• Companies with massive user bases and strong cash flow, like Google, are in a better position to invest aggressively and wait for AI payoffs.
• Valuation still matters. Even great businesses can become risky if their stock price runs too far ahead of fundamentals.

For long-term investors, Google represents a blend of classic Buffett traits—strong cash generation, a wide moat, and durable demand—with the growth potential of cutting-edge AI. That combination is rare, and it helps explain why one of the world’s most conservative investors is finally making a big bet on AI.