The tech giant deprecated its employee-built “KiroRank” ranking system after concluding it was pushing staff to inflate AI activity rather than solve genuine business problems, as major corporations across Silicon Valley begin reassessing the true value of runaway AI spending.
Amazon has shut down an employee-created artificial intelligence leaderboard that tracked token consumption across its workforce, after a senior executive issued a direct warning to staff against using merely to improve their standing on a dashboard rather than to address real business challenges.
The internal ranking tool, known as “KiroRank,” was created informally by a group of Amazon employees and had been tracking AI token usage as a proxy for productivity. The system was deprecated following concerns that it was incentivising a behaviour now widely known in Silicon Valley as “tokenmaxxing”: the practice of measuring AI-driven output by sheer volume of consumption rather than by any meaningful outcome.
What Amazon’s Senior Vice President Said About AI Misuse at the Company
The company’s discomfort with the leaderboard became public after Dave Treadwell, a senior vice president at Amazon, addressed employees directly. “Please don’t use AI just for the sake of using AI,” Treadwell told staff earlier this week. “Use AI to help you solve customer problems, to help you solve business problems, to innovate.”
An Amazon spokesperson confirmed to Business Insider that KiroRank had been “deprecated,” clarifying that the dashboard was an informal initiative and “was never intended to promote the use of AI for usage’s sake.”
Amazon Says Teams Retain Freedom to Track AI Usage Their Own Way
Despite winding down has not banned AI productivity tracking altogether. Individual teams across the organisation retain the freedom to determine how they deploy AI tools and how they measure that usage internally.
What the company is clearly distancing itself from is the notion that a higher token count signals greater efficiency or innovation. Amazon does track AI token consumption for cost-management purposes, but the spokesperson made clear the company does not encourage tokenmaxxing as a philosophy.
“We’re focused on AI adoption and sharing best practices to celebrate innovation and operational efficiency gains across the company, and we’re proud of the way our teams are embracing this technology,” the spokesperson said.
In April,was reported to be closely monitoring how engineers across the business engaged with AI tools, including monthly adoption rates, how deeply AI capabilities had been embedded into workflows, and what concrete results those deployments were generating. The focus, in other words, was on outcomes rather than raw activity.
Big Tech Rethinks AI Spending as Returns Fail to Keep Pace
Amazon’s move comes at a moment of broader recalibration across the technology industry. After years of largely unchecked enthusiasm for AI investment, some of the world’s most prominent companies are beginning to question whether the financial outlay is generating proportional productivity gains.
Uber’s chief operating officer, , recently acknowledged that the rideshare giant was not seeing productivity and other gains in line with its increased AI spending. His comments followed a moment that drew considerable attention in technology circles: Uber’s chief technology officer, Praveen Neppalli Naga, revealed that the company had already exceeded its entire annual Claude Code budget by April.
Together, these disclosures suggest that the era of free-wheeling AI investment, justified by anticipated future returns, is beginning to give way to harder scrutiny of what the spending is actually producing.
What Is Tokenmaxxing and Why Is It Falling Out of Favour
, serving as the building blocks of AI chatbots, coding assistants, and automated agents. Token usage across the industry has climbed sharply throughout 2026, driven in significant part by the proliferation of agentic AI systems that can operate autonomously, with minimal human oversight, for extended periods.
As that usage has grown, so too has the temptation to treat token consumption as a performance metric, an easy number to track, report upward, and optimise for, regardless of whether the underlying work it represents is genuinely useful.
The KiroRank episode illustrates precisely that risk. When employees are ranked by how much AI they use, the incentive structure rewards usage, not value. Amazon’s decision to shut the leaderboard down, and to say so publicly, is a signal that at least one of the largest technology companies in the world is no longer comfortable with that trade-off.