Amazon employees are inflating AI usage to top leaderboards and impress managers

Amazon employees are reportedly gaming internal AI usage metrics on MeshClaw by running unnecessary tasks to boost token counts.

The behavior appears tied to leaderboards and a desire to impress managers, despite company guidance that AI usage should not affect performance reviews.

The case shows how AI adoption metrics can be distorted by incentives, making usage data less reliable.

It also raises broader concerns about unhealthy workplace pressure and security risks as workers automate more tasks.