Amazon is closing the door on one of the internet’s oldest crowdsourcing services. According to TechCrunch AI, a notice on the Mechanical Turk website says the platform will stop accepting new customers on July 30, 2026. Amazon Web Services says it made the call after “careful consideration,” and confirmed that existing customers can keep using the service as normal.
This isn’t a full shutdown. But it’s close.
What’s actually happening
AWS isn’t pulling the plug entirely. The company told TechCrunch AI it will keep investing in “security and availability improvements” for Mechanical Turk. The catch: “we do not plan to introduce new features.” No new customers, no new development. That’s a maintenance-mode announcement, and maintenance mode is usually the last stop before the exit.
Here’s what the change means in plain terms:
- New sign-ups end July 30, 2026
- Current customers can keep running tasks for now
- No new features are coming, only security and uptime patches
- The long-term future of the service looks shaky
Why Mechanical Turk mattered
Launched in 2005, Mechanical Turk (or MTurk) was a marketplace where people got paid tiny sums to do simple tasks that machines couldn’t handle on their own. Think solving CAPTCHAs or tagging the sentiment of a sentence. It became a fixture in academic research and a lightning rod in debates over the ethics of crowdsourced micro-labor. TechCrunch AI notes it even played a small role in the early days of the Facebook-Cambridge Analytica scandal.
Starting in 2018, Amazon repositioned the service. It became a way for companies to annotate data and train neural networks through its SageMaker AI product. In other words, MTurk quietly became part of the machinery that trains modern AI.
It also earned a reputation as the hidden hand behind plenty of “AI” products that were really just people doing the work behind the curtain. Fitting, given the name. The original 18th-century Mechanical Turk was a hoax, a chess-playing “machine” with a human player hidden inside.
The irony that broke it
What stands out here is how AI itself may have hollowed out the platform. TechCrunch AI points to a 2023 analysis that found between 33% and 46% of MTurk workers were using large language models to complete their tasks.
Sit with that for a second. A platform built to supply human judgment for training AI was increasingly staffed by humans quietly outsourcing their judgment back to AI. That’s a serious problem for anyone relying on the data being genuinely human-labeled, and it raises the obvious question: if a bot does the task anyway, why keep a human in the loop?
One Reddit user quoted by TechCrunch AI argued the platform effectively died “years ago,” with workers and researchers walking away over bots and fraud. Their prediction: “Someone at Amazon is going to decide keeping the Mturk servers running is a waste of time and resources and pull the plug entirely.”
Why this matters for the AI industry
Data labeling is the unglamorous foundation of most machine learning. Every image classifier, sentiment model, and content filter learns from data that a human somewhere tagged. MTurk was one of the first places that work happened at scale, and its decline says something bigger about where the industry is heading.
Three takeaways worth holding onto:
- Human annotation is consolidating. The market has shifted toward specialized vendors like Scale AI, Surge, and others that promise vetted, higher-quality labelers instead of an open marketplace.
- AI is eating its own supply chain. When workers use LLMs to fake human labels, the training data gets contaminated, and model quality quietly suffers. Buyers now have to police for it.
- The “human behind the AI” era is ending. Fake-it-till-you-make-it products that leaned on cheap human labor have fewer places to hide.
If you rely on MTurk for research or data work, start planning your migration now. Existing access continues for the moment, but a service with no new features and no new customers is a service on borrowed time. You can find the full details at the original source.