Amazon uses AI to evaluate and sack warehouse workers

閱讀中文版本

According to the documents obtained by The Verge, Amazon is currently deploying an automated tracking and termination process. The system tracks the rates of each fulfilment centres workers' productivity and automatically issues warnings or terminations. Amazon claimed supervisors are able to override the process.

Amazon’s standards are strict. Employees (so-called Associates by the company) must be meticulous in processing each order, and therefore the company has developed a proprietary productivity metric where goals are set objectively and are based on metrics like customer demand and location.

Other than monitoring whether employees are reaching the standardised productivity benchmarks, the system also tracks time off tasks, which the company abbreviates as TOT. If workers take breaks from their duties for too long or too often, the system automatically generates warnings, and the employee can be fired after several warnings.

In Amazon's defence, retraining is part of the process to get workers up to standards , and they only change rates when more than 75% of workers at a facility are meeting goals. The bottom 5% of workers are placed on a training plan, according to the company. An appeal system is also part of the termination process.

Amazon's fulfilment centres are massive warehouses where workers track, pack, sort and shuffle each order. According to The Verge's estimation, Amazon was firing more than 10% of its staff annually, solely for productivity reasons. Amazon operates more than 75 fulfillment centres with more than 125,000 full-time employees, suggesting thousands lose their jobs with the company annually.