AI tools do not work like the software your company bought a decade ago. Many tech vendors now charge based on how much you actually use, not a flat fee. Each query eats computing power, and those costs stack up fast.
Bloomberg has reported that both Uber Technologies Inc. and Walmart Inc. put limits on employee AI use to keep expenses from ballooning. EY is trying to get ahead of the same problem.
Dan Diasio, who leads EY's global consulting AI practice, said the new office will give senior leaders "end-to-end visibility of where and how our investments in AI are driving material and tangible impact in our business."
The concern is real. The gap between the hype and the payoff is getting harder to ignore.
Get the free Always Be Buying eBook and learn the simple system for building wealth on any income
EY is not just tracking costs. It is treating its AI agents - software that can complete tasks on its own - like a workforce that needs management.
Errol Gardner, a vice chair at EY who leads global consulting, put it plainly: "You can't expect the HR or the talent function to manage an agentic workforce, and you can't expect tech to manage an agentic workforce, and you probably can't expect your finance function to manage an agentic workforce."
That is why the new agent economics leader will report to Anthony Caterino, a managing partner who oversees business enablement globally. The office is meant to keep the rollout from splintering into separate departmental efforts. Gardner said EY is also urging clients to set up similar oversight models.
The company has already started making changes. A token is the basic unit used to meter AI computing. Workers who burn through their tokens have to get approval for more.
EY is also routing around the problem. Diasio said EY is thoughtful about when it actually needs the most advanced "frontier" AI models, which cost the most to run.
Here is the angle worth watching. Public companies are spending enormous sums on AI, and their shareholders are starting to ask when the returns show up. If a firm as sophisticated as EY needs a dedicated office to keep costs under control, the odds are good that other companies are wrestling with the same thing.
When companies figure out how to make AI cheaper and more efficient, that is good news for their profit margins. When they cannot, those costs land somewhere - often in higher prices for customers or thinner earnings for investors.
The real takeaway is that the AI boom is maturing. The first phase was about who could build and buy the most impressive tools. This next phase is about who can use them without blowing up the budget. For your money, the companies and products that solve that puzzle are the ones worth paying attention to.
Download the free Always Be Buying eBook and start putting your money to work today