Amazon employees are doing fake jobs because they are forced to use a lot of AI and show it

The AI ​​race for business is slowly starting to feel less like innovation and more like performance art. Companies are desperate for workers to “embrace AI,” workers are desperate to be managed off their backs, and somewhere in between, everyone is now automating tasks no one needed automated in the first place.

According to a new report by the Financial Times, Amazon employees are using the company’s internal AI tool called “MeshClaw” to perform redundant tasks just to boost their AI usage scores and appear more in line with the company’s growing AI-first culture. In context, Amazon’s MeshClaw can trigger code submissions, check emails, and interact with apps like Slack, according to people familiar with the matter.

Amazon’s internal AI push has reportedly turned into theater

A report says that Amazon recently launched an internal target that encourages more than 80% of developers to use AI tools on a weekly basis. That pressure has reportedly pushed some employees into outsourcing low-value or completely unnecessary work to AI agents to climb internal leaderboards and display acquisition metrics.

And honestly, this sounds like the most predictable outcome imaginable. When companies began to integrate employee performance and visibility into the adoption of AI, it was inevitable that some employees would begin to prepare for an “AI-friendly appearance” rather than productivity.

Amazon is not alone here either. As reported by Wired, Meta was reportedly facing internal backlash from employees who were unhappy about the heavy-handed AI training processes, including mouse tracking and monitoring systems tied to AI job development. Meanwhile, another recent report suggested that even Meta employees are struggling to integrate AI into daily work despite leadership pushing it internally.

The funny thing is that AI is more expensive than real people

This is where the whole AI gold rush starts to look ridiculous. Recent Axios reports have already suggested that, in a few cases, enterprise AI systems are more expensive than simply paying for human workers, especially when token costs, infrastructure, and scaling costs are factored in.

And somehow, despite all that, companies are still laying off workers to chase AI adoption metrics while many AI firms continue to sell products at a loss just to grab early market share. That’s the part that no one seems ready to talk about yet. Currently, these tools are “cheap” because the industry is still supporting growth. But if businesses become completely dependent on AI workflows and human jobs already disappear, those pricing models could change very quickly.

Honestly, this doesn’t feel like a productivity revolution anymore. It feels like the tech industry is racing to the top of another expensive bubble while real jobs are quietly disappearing in the background.