Consulting giant Bain sounds the alarm: The global AI industry needs to achieve $6 trillion in annual revenue to sustain the “cash burn” of data centers

According to Zhitong Finance APP, global consulting giant Bain & Company stated that by 2031, the global artificial intelligence (AI) industry would need to generate $6 trillion in annual revenue to justify the massive worldwide capital investment in building data centers.

Bain & Company noted in its annual global technology report released on Tuesday that existing consumer and enterprise AI services may contribute up to $1.8 trillion of that total, which means $4.2 trillion in new revenue still needs to be created. Bain said this revenue gap will likely come from sectors that are still in their infancy, such as autonomous machines and robotics, as well as emerging fields like drug development, mental health, and energy production.

David Crawford, the primary author of the report and Head of Bain's Global Technology, Media, and Telecommunications division, stated: "This industry needs a wave of innovation on a scale far surpassing what mobile internet and cloud computing unleashed. The construction of AI infrastructure is moving well ahead of the demand curve, and sustaining this pace of funding would require increasing the global annual GDP growth rate by about 1 percentage point."

Bain pointed out that although current discussions are mainly focused on employee productivity, from the perspective of the economic benefits of AI infrastructure, trillions of dollars in new revenue will be needed in addition to productivity improvements.

In addition, Bain expects data center expenditures to reach $5 trillion to $6.5 trillion by 2030, adding at least 150 gigawatts of new capacity, which will further intensify national pressures on energy resources. The company said that by 2031, annual spending on AI infrastructure—including data centers, computing power, as well as accelerators and memory chip upgrades—could reach as high as $1.5 trillion.

Data center developers are already facing shortages of transformers, water, and electricity supplies, as well as strong local community opposition. In the US alone, $68 billion worth of data center projects have been blocked or delayed in the second quarter.

Bain’s report highlights that for current AI development speeds to be sustainable, there are still many obstacles ahead. Companies like Microsoft, Google, Amazon, Meta Platforms, and Oracle are investing trillions of dollars in building data centers to meet the surging demand for AI computing resources. The scale and cost of data centers are doubling approximately every 12 to 16 months, partly due to soaring prices for chips, network equipment, and other components from companies like Nvidia and SK Hynix.

As AI service providers have yet to realize clear returns, this report is released amid rising debate about the returns on AI investments. Critics worry that a growing network of mutual dependence is forming between tech manufacturers and AI developers, fueling optimistic market expectations that in turn require more and more capital input.

It is worth noting that the possibility that huge AI capital expenditures may not result in sustainable returns is also a crucial reason why Michael Burry continues to take a bearish stance on AI-related trades. The investor, renowned for heavily shorting the US housing market ahead of the global financial crisis, recently stated that the AI bubble could burst earlier than initially thought. He is now converting his heavily shorted AI positions into put options, enabling him to acquire more cost-efficient leverage within a shorter time frame.

In an investment newsletter released on Monday, Burry wrote: "Essentially, I'm moving the timetable up. I therefore want my short positions to have more leverage. When the timing window is tighter, leverage is easier to handle. For leverage, nothing fits better than options—specifically, puts; as volatility indicators like the VIX remain abnormally low, these options are relatively cheap." He revealed that part of the position adjustment is to reduce tax liability, but the main reason is his belief that "the AI bubble might burst sooner than expected." The new put option positions mean he is betting that the AI trade could reverse before next summer.