An OpenAI executive revealed in court that the company expects to spend around $50 billion on computing power before the end of 2026. The statement was made during a legal hearing involving the company and Elon Musk. This staggering figure highlights the immense costs involved in developing advanced artificial intelligence models today.

The Scale of OpenAI’s Computing Expenses

The $50 billion figure was shared by Greg Brockman, OpenAI’s cofounder and president, and was previously reported by Bloomberg. It underscores how much money is pouring into AI development, especially for training large models like ChatGPT. Despite its popularity, OpenAI has yet to turn a profit and struggles to meet its revenue goals.

The company’s heavy spending is largely tied to leasing vast amounts of compute resources from major cloud providers. OpenAI’s partnerships with companies like Microsoft, Amazon, Nvidia, and SoftBank involve significant investments that are often contingent on the company using large-scale cloud services. These deals include commitments to rent thousands of gigawatts of AI training hardware, such as Amazon’s Trainium accelerators and Nvidia’s GPUs, to run its models effectively.

Investments, Rebates, and the Question of Profitability

Many of the investments announced by OpenAI’s partners are not straightforward donations but rather complex financial arrangements. For example, Amazon’s multi-billion dollar commitment includes requirements to deploy OpenAI’s models on AWS infrastructure, effectively providing a discount or rebate. Nvidia’s investment also involves deploying thousands of GPUs, with costs reaching hundreds of billions of dollars in the broader hardware expense landscape.

This raises questions about whether OpenAI can truly burn through $50 billion in a single year or if these numbers are more about maintaining a public image of rapid growth and momentum. Some analysts suggest that the large sums are partly subsidized through discounted cloud services rather than outright spending. OpenAI has not publicly clarified how much of this spending is direct costs versus subsidized deals.

The company’s strategy seems focused on scaling up compute resources as quickly as possible to stay ahead in the AI race. However, with no clear path to profitability yet, critics wonder whether these massive investments are sustainable long-term. The industry continues to watch closely as OpenAI and its partners push the boundaries of AI hardware and software development.

Inspired by

• https://go.theregister.com/feed/www.theregister.com/2026/05/05/openai_50b_compute/

Sources

• bloomberg.com
• wsj.com