Artificial intelligence is increasingly gaining ground in large-scale disputes and investments. Global banks are discussing lending US$38 billion to finance a new wave of data centers for OpenAI systems. This news reflects the current global context, where the competition for computing power is becoming a determining factor in the technology sector. For everyday consumers, the effects of this can already be seen in tools used for study, work, and even creating art.
OpenAI seeks to secure the computational foundation that will support its next era
The logic behind these investments is actually quite simple. Without energy and specialized chips, there is no necessary infrastructure for artificial intelligence to support an expanding user base. OpenAI, creator of ChatGPT and other generative models, is a pioneer in this area and understands this need. Recently, it finalized one of the largest digital infrastructure contracts in history: US$300 billion in computing power with Oracle for the next five years. This agreement is part of the so-called Project Stargate, a plan aimed at building next-generation data centers.
The physical scale of the project redefines the limits of global technological infrastructure
The financial investment is quite significant, but it accurately reflects the physical scale of the project, which is also impressive. OpenAI and its partners want to install approximately 4.5 gigawatts of energy capacity, which can be compared to the output of two Hoover Dams and exceeds the electricity consumption of four million American homes.
Meanwhile, other tech giants like Google, Amazon, Meta, and Microsoft also plan to invest around US$300 billion this year in AI-focused data centers. OpenAI alone already moves amounts close to those of its competitors combined.
Global banks assess the risks and promises of this technological race
The new $38 billion funding request is being reviewed by major international financial institutions. A sum of this magnitude represents a bet by financial institutions on the potential of AI to drive the global economy.
Like any investment, this one is not without risk. OpenAI is not yet profitable due to the scale of financial commitments it undertakes. The leap into data centers dependent on “hundreds of thousands” of Nvidia chips requires long-term outlays and the confidence that consumers, businesses, and governments will continue to adopt these systems. The company has also partnered on infrastructure projects with Oracle, SoftBank, and manufacturers such as Nvidia, AMD, and Broadcom.
The relationship with partners changes as OpenAI itself evolves. For years, Microsoft was the company’s main cloud provider. But, as with Amazon-Anthropic, the landscape is rapidly reconfiguring itself. OpenAI now operates with greater strategic autonomy, seeking redundancy, that is, with a “plan B” so that if one network fails, another takes over. For banks, this is essential for a more risk-free proposition.
The impact on ordinary users and the global technology market
The competition for computing power is not just a power struggle, but also a matter of necessity stemming from the high concrete demand of millions of people. Systems like ChatGPT need to answer billions of questions per day, simultaneously. And when new models emerge, such as more creative, multimodal versions with advanced memory, energy consumption grows at an almost exponential rate.
The proposed funding would serve to accelerate this cycle and ensure greater stability for the high volume of work demanded by the platforms. These would be data centers dedicated exclusively to AI, focused on reducing latency and, above all, providing space for new algorithms to be tested without disrupting current operations.
If this funding level is ultimately approved by financial institutions, OpenAI will be in a prominent position relative to its competitors, securing resources that ensure the experience of billions of users and guaranteeing the testing of new ventures in parallel.
