OpenAI and its flagship product ChatGPT serve as the foundational drivers for the current AI economic bubble, according to analyst Ed Zitron. Since Microsoft’s initial $1 billion investment in 2019 and the subsequent launch of its GPU supercomputer, OpenAI has become the primary justification for over $1 trillion in global capital expenditures. While other companies like Anthropic have received significant funding, their existence is largely predicated on hyperscalers attempting to recreate or compete with OpenAI’s success. OpenAI currently accounts for more than 50% of global AI compute infrastructure demand, with plans to burn over $852 billion by 2030, a figure sustained by a recent $122 billion funding round.

The bubble is characterized by a lack of tangible return on investment, including revenue, profitability, or verified productivity gains. Instead, the industry relies on the rapid user growth of ChatGPT—which reached 100 million weekly active users by late 2023—to validate massive infrastructure investments. Hyperscalers such as Microsoft, Google, and Amazon have bankrolled these AI labs, providing credits and hardware that allow companies to operate without traditional venture capital constraints. This cycle creates a monoculture where investors assume the model established by OpenAI can be replicated infinitely, ignoring that the primary consumers of AI compute are limited to a small group of hyperscalers and their incubated partners.

Data center expansion and debt deals, such as the $178.5 billion in US-based data center debt recorded in 2025, are driven almost entirely by the insatiable compute requirements of OpenAI and Anthropic. Because few other companies have successfully scaled AI clusters, hyperscalers have effectively created a closed system where they fund the infrastructure, the startups purchase the capacity, and the resulting revenue is used to justify further spending. This systemic dependency suggests that OpenAI’s potential failure would serve as a watershed moment for the industry, potentially ending the current epoch of capital misallocation and forcing a broader reassessment of AI profitability and demand.

OpenAI and its flagship product ChatGPT serve as the foundational drivers for the current AI economic bubble, according to analyst Ed Zitron. Since Microsoft’s initial $1 billion investment in 2019 and the subsequent launch of its GPU supercomputer, OpenAI has become the primary justification for over $1 trillion in global capital expenditures. While other companies like Anthropic have received significant funding, their existence is largely predicated on hyperscalers attempting to recreate or compete with OpenAI’s success. OpenAI currently accounts for more than 50% of global AI compute infrastructure demand, with plans to burn over $852 billion by 2030, a figure sustained by a recent $122 billion funding round.

The bubble is characterized by a lack of tangible return on investment, including revenue, profitability, or verified productivity gains. Instead, the industry relies on the rapid user growth of ChatGPT—which reached 100 million weekly active users by late 2023—to validate massive infrastructure investments. Hyperscalers such as Microsoft, Google, and Amazon have bankrolled these AI labs, providing credits and hardware that allow companies to operate without traditional venture capital constraints. This cycle creates a monoculture where investors assume the model established by OpenAI can be replicated infinitely, ignoring that the primary consumers of AI compute are limited to a small group of hyperscalers and their incubated partners.

Data center expansion and debt deals, such as the $178.5 billion in US-based data center debt recorded in 2025, are driven almost entirely by the insatiable compute requirements of OpenAI and Anthropic. Because few other companies have successfully scaled AI clusters, hyperscalers have effectively created a closed system where they fund the infrastructure, the startups purchase the capacity, and the resulting revenue is used to justify further spending. This systemic dependency suggests that OpenAI’s potential failure would serve as a watershed moment for the industry, potentially ending the current epoch of capital misallocation and forcing a broader reassessment of AI profitability and demand.