Amazon has told its software engineers to stop padding their AI usage just to look productive, according to an internal memo circulated in September and posted publicly on the company’s website this week. The practice of extreme AI usage, popularly known as “tokenmaxxing,” has been encouraged by tech giants, but now that they face huge usage bills, they seem to be cutting back.
CNBC reports that the Amazon memo discourages what is known as “tokenmaxxing,” running up AI token consumption to look more active with the technology. It also redraws how Amazon defines the roles of software development engineers and managers now that AI tools sit inside daily coding work. “We’re just seeing the beginning of AI’s effect on software engineering,” the memo states.
Companies across the tech industry are wrestling with the rising cost of AI tools and struggling to measure whether the technology actually makes workers more productive. Some firms have capped how many tokens employees can burn through; others have shifted tasks to cheaper, open-source models to hold down costs — including Chinese alternatives. Amazon has not gone that route. A company spokesperson said Amazon has not instituted caps on employee token usage.
Amazon has run into this problem before. In May 2026, the company shut down an internal leaderboard after discovering employees were gaming their AI usage numbers to climb the rankings, a direct example of the tokenmaxxing the new memo warns against.
The memo singles out three measurements: token usage, AI tool adoption rate, and the volume of code changes and reviews, as “deeply flawed” ways to gauge how AI is being used. Leaning on them, it warns, can drive up costs and add unnecessary complexity to engineering work. Amazon isn’t telling employees to throw those numbers out entirely; the memo says they should be weighed “with care and judgment” alongside other data. “We believe that it is best to measure, and be measured by, long-term outcomes,” the memo states.
The document also leaves room for engineers to say AI simply isn’t helping with a particular task. “If your work does not benefit from current AI tooling, that is a finding, not a failing,” it reads, adding that “an engineer who can explain where AI does not yet help their work is demonstrating the judgement we are asking for.”
Even with those caveats, the memo treats AI as central to where software engineering is heading, even if the details are still being worked out. “While we have a vision for what the future looks like, there’s a lot that’s still emerging,” the memo says. “We are certain that whatever future best practices are, AI will be at the center of software engineering practice.”
Inside Amazon Web Services, the company is already leaning on a different yardstick. Marc Brooker, a vice president and distinguished engineer at AWS, told CNBC that Amazon looks at “feature velocity,” tracking “how many launches are we doing, how many successful changes are we making in production,” to judge AI’s effect on engineer output. Brooker described Amazon’s overall approach to AI spending as a return-on-investment calculation. “Broadly, we think about this as a return on investment,” he said. “If there is an important problem to solve for our customers, important problem to solve for our business, we’re going to invest tools, time, tokens, human effort, etc. into those.”
He also pointed to the tension between discipline and experimentation. “What we don’t want to see is wasted effort where we are going and spending those tokens solving problems that don’t matter and sometimes it’s hard to know, so we do need to have some appetite for experimentation,” Brooker said.
The memo lands against the backdrop of broader predictions from Amazon’s leadership about AI’s effect on the workforce. Chief Executive Andy Jassy has said AI will “change the scope and speed” of everyday work, and has previously said AI-driven automation could shrink Amazon’s corporate workforce in the coming years, while also creating new kinds of jobs.
Companies ranging in size from Amazon to garage startups are struggling to determine the best way to harness the potential of AI. Breitbart News social media director Wynton Hall has written his New York Times bestseller Code Red: The Left, the Right, China, and the Race to Control AI to serve as the definitive guide on how the MAGA movement can create positions on AI that benefit humanity without handing control of our nation to the leftists of Silicon Valley or allowing the Chinese to take over the world.
Read more at CNBC here.
Lucas Nolan is a reporter for Breitbart News covering issues of AI, free speech, and online censorship.
