In the fight for artificial intelligence, the biggest IT corporations are changing from software-centric businesses to major global infrastructure owners. There has been a claimed 140 percent increase in the value of Amazon, Alphabet, Microsoft, and Meta’s combined property, plant, and equipment over the last three years, bringing the total to $1.46 trillion. The four businesses are now able to compete with, and in some cases even surpass, global oil giants in terms of physical assets, thanks to their aggressive spending on AI data centers, servers, and networking equipment. According to the article, the four US IT businesses’ combined PP&E assets had grown by 48% year-on-year by the end of June, according to the figures cited. Property, plant, and equipment assets at $538.7 billion are nearly twice as high as they were three years ago, putting Amazon at the top of the group. Prior to this year, Saudi Aramco was the biggest non-financial company in the world in terms of physical assets; however, the corporation has already overtaken them.
Both Alphabet and Microsoft now have more than $330 billion in PP&E assets, which is more than major oil companies like PetroChina and Exxon Mobil. Meta has grown substantially, despite being the smallest of the four. If the claim is to be believed, the present worth of its physical assets is higher than Toyota Motor’s. Investments in AI infrastructure are largely responsible for the rapid growth in assets. More than 70% of Alphabet’s PPE consists of technical infrastructure, which includes data center land and buildings, computers, and networking gear. All four of these businesses are steadily and rapidly increasing their AI capabilities. They may spend as much as $760 billion on capital expenditures in 2026, up over 85% from the previous year. The spending highlights how AI is reshaping the business models of top tech companies. Companies that formerly relied heavily on software are increasingly investing more in hardware to meet the growing need for AI computing. Investments in artificial intelligence are booming, yet companies’ balance sheets do not yet represent the full extent of these commitments.
Around $2.3 trillion in off-balance-sheet obligations, including leasing commitments and long-term equipment purchase agreements, were held by the four firms as of the end of June. This amount was more than four times higher than the same period last year, as stated in the study. These liabilities, sometimes called “hidden debt,” are expected to transform into real assets in due course. Over the past year, Alphabet’s off-balance-sheet commitments have increased by nine times, while Meta’s have increased by eight times. A different research indicated that the aggregated undisclosed debt of Alphabet, Microsoft, Amazon, Meta, and Oracle has climbed to almost $1.65 trillion in the last four years, almost eight times higher than their disclosed debt of about $1.35 trillion. Meta has the highest amount of projected concealed debt among the five businesses, nearing $420 billion—nearly three times its disclosed borrowings. One of the most notable increases has occurred at Oracle. By the end of May, its hidden liabilities had grown to around $273.3 billion, driven by long-term leasing commitments related to its Stargate AI data center project with OpenAI.
The investment in AI infrastructure is expected to pay off in the long run thru cloud computing and AI services, but it will also drive up operational expenses. Companies will have to account for higher depreciation expenses when they build out their server and data center infrastructure, since AI gear typically only lasts about five years. According to the research, Amazon, Alphabet, Microsoft, and Meta incurred a combined depreciation charge of $44.5 billion during the April-June quarter, which accounted for over one-third of their total operating profit. The first drop in operational profit for Meta in three years was attributable, in part, to higher depreciation expenditures. According to market projections, the four businesses’ yearly depreciation charges might reach around $360 billion by 2028, which is nearly twice the level expected for 2026.