Amazon Mechanical Turk Freezes New Sign-Ups (2026): How AI Killed the Original Crowdsourcing Platform

Amazon Mechanical Turk (MTurk), the pioneering crowdsourcing marketplace launched in 2005, will close to new customers on July 30, 2026, as confirmed by an official banner on the MTurk homepage. The decision, announced on July 3–5, 2026, places the platform into "maintenance mode"—AWS's designation for a service that is no longer actively developed and is effectively slated for eventual retirement.

Why Is Amazon Shutting Down Mechanical Turk to New Customers?

The primary driver is the rapid advancement of artificial intelligence. According to a TechCrunch report, AWS stated that the move follows "careful consideration" of the platform's declining relevance. Generative AI models, such as large language models (LLMs), can now perform the simple, repetitive tasks—like image tagging, data verification, and content moderation—that once required human workers on MTurk. The irony is thick: MTurk was originally dubbed "artificial artificial intelligence" because humans hid behind the scenes to do tasks AI couldn't. Now, real AI has made those humans redundant.

A Gizmodo article highlights a 2023 study that found between 33% and 46% of MTurk workers were already using LLMs to complete their tasks. This self-defeating cycle undermined the very value proposition of verified human annotations: if workers themselves use AI, why pay humans at all? Moreover, researchers have increasingly abandoned MTurk due to bots and fraud, further eroding trust.

How Did AI Accelerate MTurk's Decline?

The Register points out that Amazon's own internal AI services, such as SageMaker Ground Truth, have cannibalized MTurk's data-labeling business. As AWS invested heavily in generative AI infrastructure—Amazon just returned to the bond market to raise $25 billion for AI buildout, per Reuters—it made little sense to maintain a manual labor platform that could not keep up. The Register's piece, titled "Amazon's Mechanical Turk to stop accepting new customers – and not even AI can save it," notes that MTurk's attempted pivot to a data-labeling solution for AI was too little, too late.

Meanwhile, Mercor's $2B Run Rate Signals a New Model

As MTurk fades, a startup called Mercor has rocketed to a $2 billion gross run rate, as reported by RuntimeWire. Mercor is an expert-data marketplace that connects workers with advanced skills—often possessing degrees or specialized knowledge—to AI labs needing high-quality training data. Unlike MTurk's general microtask pool, Mercor focuses on tasks that still require genuine human expertise, such as coding and scientific reasoning. The company pays 60–70% of revenue to contractors, and its growth doubled in just a few months. This suggests that the future of human labor in AI is not in millions of anonymous microtaskers, but in a smaller, more skilled workforce.

What Does Maintenance Mode Mean for Existing MTurk Users?

Existing users—both requesters (the companies posting tasks) and workers—can continue using the platform as before, at least for now. AWS has stated it will maintain security and availability, but no new features will be developed. This effectively puts MTurk on "life support." The official banner on mturk.com reads: "New customers are no longer being accepted as of July 30, 2026. Existing users are not affected." However, the lack of updates means eventual retirement is likely.

Table: MTurk Timeline and Key Data Points

Event / Metric	Date / Value	Source
MTurk launch	2005	Industry knowledge
New sign-ups stop	July 30, 2026	MTurk homepage
% workers using LLMs	33–46% (2023 study)	Gizmodo
Mercor gross run rate	$2B (June 2026)	RuntimeWire
AWS maintenance mode status	Indefinite, no new features	The Register

What Does the MTurk Sunset Mean for the Crowdsourcing Industry?

The end of new MTurk registrations signals a broader shift. For years, MTurk was synonymous with crowdsourced microtasks, powering everything from academic surveys to startup data pipelines. But its low pay (often below minimum wage) and ethical controversies—such as involvement in the Facebook-Cambridge Analytica scandal—drew increasing criticism. As AI improves, the economic logic of paying humans pennies for tasks that a model can do instantly vanishes.

However, human labor is not entirely obsolete. Highly specialized tasks—like legal document review, medical data labeling, or creative work—still require human discernment. Mercor's rise shows that companies are willing to pay premium rates for verified experts. Meanwhile, Amazon has its own competing annotation services through AWS, and third-party providers offer alternative solutions.

What Should Workers and Requesters Do?

Workers who rely on MTurk for income should explore alternatives such as Prolific, Clickworker, or specialized platforms like Mercor for AI training. Requesters should migrate their workflows to AWS's own annotation tools or to API-based AI services. Given that MTurk will receive no new features, relying on it for future projects carries risk.

Conclusion

Amazon Mechanical Turk's closure to new customers is a watershed moment for the gig economy and AI industry. It marks the end of a 21-year experiment in "human-as-a-service" labor, killed by the very technology it was designed to train. As AI continues to mature, the human-in-the-loop model will not disappear but will evolve toward higher-skill, higher-pay tasks. The era of cheap, anonymous microtasking is over.

For further analysis, read the Hacker News discussion on these paired developments.