The Companies Dominating AI Investment

Last October, Mark Zuckerberg made a statement during an earnings call that should have received more attention. He described Meta’s place in the AI race, stating that the company was running its suite of applications “in a compute-starved state.” Meta, a business that made tens of billions of dollars from advertising last year, is experiencing a computer power shortage. Meta’s 2025 budget had already increased to between $70 billion and $72 billion, and the company indicated that 2026 would be “notably larger.”

This comment was intended to justify a massive increase in capital spending. As I listened to it, the scope of what was being described seemed almost dizzying. This isn’t a product line wager. It’s a complete reorganization of a business centered on unfinished infrastructure.

Key Facts: Companies Dominating AI Investment (2025–2026)

Meta is not by herself. Alphabet increased its own capital expenditure forecast for 2025 to between $91 billion and $93 billion the same week those earnings were released. This is nearly twice as much as the company spent in 2024. In just three months, Microsoft revealed quarterly data center spending of $34.9 billion, which shocked analysts and was much higher than the $24 billion spent in the previous quarter. During that time, all three businesses performed better than the S&P 500 as a whole.

It appears that investors are interpreting the expenditures as a commitment rather than a sign of recklessness—the kind of long-term infrastructure investment that often appears costly until it doesn’t. Well, that’s the wager. Wall Street is closely monitoring the revenue lines for indications that the data centers are producing returns commensurate with the capital invested in them, as it is still genuinely unclear whether it will pay off.

In all of this, Nvidia is in an intriguing position. The company manufactures the chips that enable AI, providing the processing power that Meta, Alphabet, and Microsoft are purchasing by the truckload, rather than creating the applications that use AI. Nvidia’s next-generation Rubin platform, which includes six new chip designs, was introduced in January 2026 with the goal of enabling users to create and execute the biggest AI workloads at a reduced cost. Argus analysts set a price target of $220 for the stock, which closed at $186 on March 11.

Nvidia’s stance is based on nearly traditional infrastructure economics, which states that during a gold rush, those who sell equipment frequently outperform those who dig. The announcement of Rubin indicates that Nvidia, which has been that company for the current AI moment, plans to remain there as the computational demands of the upcoming wave of AI models increase.

Because Microsoft’s investment strategy is multifaceted, it merits careful examination. In addition to integrating ChatGPT into its Bing search engine and merging its diverse AI tools into a single product known as Microsoft Copilot, the company has invested over $13 billion in OpenAI. Additionally, it has spent $17.5 billion over the course of four years to develop cloud and AI infrastructure in India, a move that resembles both product strategy and geopolitical positioning.

As businesses all over the world build their AI workflows on top of Microsoft’s infrastructure rather than creating their own, Azure, Microsoft’s cloud computing division, has been expanding. Satya Nadella has consistently framed the situation as a “massive opportunity,” which is real and requires spending. Profits for the quarter increased by 12% to $27.7 billion, indicating that the present business is paying for the future wagers.

However, in the larger picture of investing, there is something worth considering. The big tech companies, the Alphabets, Metas, and Microsofts of the world, are nearly the sole focus of the conventional narrative regarding AI spending. However, a thorough examination of venture capital data reveals a more nuanced picture. The food and beverage sector increased its investment in AI by over 131,000 percent, from about $60,000 in 2012 to $76 million in 2025. AI has been used in warehouse operations for maintenance and quality control by PepsiCo, a company that most people do not associate with cutting-edge technology.

During the same time period, government and defense AI investment increased from $32 million to $1.6 billion. Financial services increased to almost $12 billion from $266 million. It turns out that the AI investment story is far more widespread than the headlines indicate, occurring covertly and without press releases in soda factories, defense procurement offices, and insurance underwriting departments.

As all of this builds up, it’s difficult to ignore the fact that the investment is now self-reinforcing in a way that makes it challenging to slow down. Senior US economist Aditya Bhave of Bank of America stated that AI-related business investment is one of the two factors supporting GDP growth in the US over the last few months, along with consumer spending.

The stakes are significantly increased by that framing. The story of which company creates the best model is no longer limited to the technology sector. It’s an economic tale about how capital is distributed throughout the modern economy’s whole productive base and what happens to growth if those distributions fail to produce the returns that the spending suggests. The businesses making the biggest wagers appear certain that they are correct. The rest of the world is keeping a close eye on things to find out.