Amazon to stop accepting new Mechanical Turk customers

Amazon Web Services (AWS), the division of Amazon managing the service, announced it will stop accepting new customers for its Mechanical Turk crowdsourcing service on July 30, 2026. AWS stated that the decision was made after careful consideration. While the platform is closing to new sign-ups, AWS confirmed that existing customers can continue to use the service as normal. However, the company does not plan to introduce new features. According to AWS, “AWS continues to invest in security and availability improvements for Mechanical Turk, but we do not plan to introduce new features.” This decision effectively places the service on life support, continuing basic operations without future development.

First launched in 2005, Mechanical Turk is a crowdsourcing marketplace designed to let users perform simple tasks. Over time, the platform became a tool for data annotation, which involves labeling data to train machine learning models. In 2018, Amazon began billing the marketplace as a way to annotate data for SageMaker AI, Amazon’s service for building, training, and deploying machine learning models. Despite its utility, the platform has also been characterized as the hidden enabler for companies taking a fake-it-till-you-make-it approach to AI, where human labor is used to power products marketed as automated. The platform was also mentioned in the context of the Facebook and Cambridge Analytica scandal.

In recent years, the platform’s relevance has waned amid concerns about bot activity and fraud. A 2023 analysis found that between 33% and 46% of workers on the platform were using large language models to complete tasks. This shift has raised questions about the reliability of data annotated on the platform. On community platforms like Reddit, users noted that workers and researchers have been abandoning the service. One Reddit user predicted that someone at Amazon will eventually decide that keeping the servers running is a waste of time and resources and pull the plug entirely, suggesting the platform died years ago.

Why it matters

Amazon’s decision signals a decline for a platform that was once a cornerstone of data annotation for AI, highlighting the shifting landscape of human-in-the-loop labor. As artificial intelligence models evolve, the reliance on crowdsourced human networks for basic data labeling is giving way to automated training methods.