Has the rush to adopt artificial intelligence across businesses—and the widespread layoffs that followed—actually produced the major savings the technology industry expected?

According to a new global survey by KPMG, one of the world's four largest accounting and consulting firms, a growing number of multinational corporate executives are now experiencing "AI bill shock" as invoices for artificial intelligence services continue to climb.

It appears that while artificial intelligence companies previously subsidized the cost of their large language models through contracts with fixed pricing, the enormous computing power now required to run increasingly sophisticated AI systems in massive data centers is reshaping the technology sector and driving significant price increases.

Only a few months ago, chief executives were encouraging—or even requiring—their employees to use artificial intelligence as much as possible for tasks such as software development, a phenomenon that became known as "token-maxxing."

Amazon even began measuring employees according to the number of AI "tokens" they consumed, as though they were competing in a video game, before discovering that some employees were launching artificial intelligence agents to perform unnecessary tasks simply to improve their rankings. Meta also encouraged greater use of AI tools by incorporating AI usage into employee performance evaluations.

Now, many organizations are beginning to hesitate. After employees became heavily dependent on AI-powered coding tools, companies are discovering that the cost of operating those systems is becoming increasingly difficult to control.

The KPMG report, first reported by the British technology news site The Register, surveyed two thousand one hundred and forty-five senior executives across twenty countries. It found that twenty-nine percent were surprised by the rising costs associated with artificial intelligence.

In other words, nearly one-third of senior executives appear to have been swept up in the excitement surrounding artificial intelligence without first developing a clear strategy for using it efficiently and economically—a reality becoming increasingly apparent now that, as the saying goes, "the meter is running."

A manager at Nvidia recently acknowledged that his research team now spends more on artificial intelligence than on employee salaries themselves. Axios also reported that an unidentified company spent five hundred million dollars in a single month on Claude usage fees after failing to place any limits on employees' licenses. Another recent study found that businesses described as the "most AI-dependent" now spend approximately seven thousand five hundred dollars per employee every month on artificial intelligence.

The report's findings reinforce concerns that a significant number of corporate leaders have treated artificial intelligence primarily as a tool for replacing existing resources—including employees—without sufficient planning, without understanding its full economic implications, and while remaining disconnected from the practical realities of implementation.

According to The Register, many executives' limited understanding of what has become known as the "economics of artificial intelligence"—a field that remains relatively new and still lacks extensive practical experience—has emerged as one of the biggest obstacles to successful deployment in the workplace.

"As usage-based pricing models become more common, rather than pricing fixed in advance, many organizations are only now beginning to develop the capabilities needed to forecast, monitor and manage AI spending effectively," the report's authors wrote.

The bad news, the publication argues, is that this situation has created a global financial environment that, by some measures, appears even more fragile than the period preceding the Great Depression.

Artificial intelligence—or, more precisely, the mythology surrounding it—is also increasingly being used as a disciplinary tool in workplaces. Fearing that AI could replace them, many employees are becoming less willing to negotiate over salaries and benefits. At the same time, many managers are using artificial intelligence not only to justify large-scale layoffs but also as a means of discouraging employees from challenging management decisions.