Some companies are encouraging employees to burn through as much AI as possible, and even ranking them by how many AI tokens they consume.

Welcome to the strange world of “tokenmaxxing.”

Tokens are essentially the currency of generative AI. They are the small chunks of text that AI models process when you enter a prompt or receive an answer, and companies such as OpenAI, Anthropic and Google typically charge businesses according to how many they use.

That makes token consumption useful for tracking AI costs. The trouble starts when businesses treat it as a measure of AI adoption, productivity or enthusiasm.

This has led to employees deliberately maximizing their token usage to climb leaderboards, hit targets or prove that they are embracing AI.

And as businesses pour billions into AI, that seemingly harmless metric could become a surprisingly expensive mistake.

Although its exact origin is unclear, the first time many people heard the term was likely to have been in reporting of Meta’s Claudeonomics leaderboard, built by an employee to rank staff by their token usage.

With workers facing pressure from above to show they were using AI to drive productivity, or at the very least just using AI, the leaderboard showed the top 250 token users. For a short while, at least, until it became public knowledge and was quickly deleted.

This could be seen as a private joke among colleagues. But it was reflective of a culture where the easily captured metric of token use was being leaned on too heavily. Particularly when it wasn’t backed with evidence that it was linked to productivity, customer experience improvement or revenue growth.

Meta was by no means the only tech giant affected. Reports emerged that workers at Amazon had worked out how to game their own leaderboard in order to appear keener adopters of AI than they actually were.

Since then, the idea has spread well beyond Silicon Valley. Business Insider reports that employees were sorted into light or heavy user categories according to their token use at JP Morgan, as well as at Disney (although Disney later clarified that it wanted to move faster without tokenmaxxing ).

Nvidia CEO Jensen Huang lent his support to the trend too, saying he would be “deeply alarmed” if a $500,000 engineer didn’t consume at least $250,000 in tokens in a year. However, he would say that his company is one of the ultimate beneficiaries.

This isn’t something that’s only happening at larger companies, either, though in the world of small business it doesn’t tend to be done so theatrically.

But I’ve heard anecdotal tales of companies telling workers to use AI as much as possible, to justify terms like “AI-first” and AI-native” in their marketing. And including AI proficiency in job descriptions where it might not be obviously necessary, so it can be used as a criterion for assessing their performance.

Over the year or so since it emerged, the term has gone from describing a strategy to primarily a warning against inefficient use of AI. Here’s why…

AI compute power is expensive, and this is particularly true when businesses start using AI agents to carry out autonomous work. Estimates put the cost of deploying agents at 10 to 100 times the cost of using regular AI chatbots.

Of course, 2026 was all about agents, so many companies found themselves burning tokens at an incredible rate. Prominent examples include Uber, which is reported to have spent its entire annual AI budget in just four months of this year.

This makes tokenmaxxing, particularly when it’s liable to be gamed, horrendously expensive for a practice that primarily maximizes effort, not results.

There are security risks too. Every agent or chatbot instance spun up creates potential security issues, like permissioning and authentication liabilities. Pointless AI work done just to climb leaderboards just increases the likelihood of something that’s been implemented insecurely causing problems.

My recommendation is to instead focus on “ valuemaxxing ”. This involves encouraging teams that use AI to report and be accountable for progress. It involves measuring impact on metrics that AI is supposed to bump up, by reporting on tasks completed, time saved and customer feedback improved.

Today, businesses are keen to show that they’re using AI, and counting tokens seems like the easiest way to do that.

Showing that it’s creating value is much harder. But it’s a bridge they’ll have to cross sooner rather than later, as shareholders, investors and customers start to demand results.