Amazon to Stop Accepting New Mechanical Turk Customers From July 30

Amazon is closing the door to new customers for Mechanical Turk, the long-running crowdsourcing marketplace that helped popularize online microtasking nearly two decades ago.

The change will take effect on July 30, 2026. Existing customers will still be able to use the service, but new customer access will no longer be available. Amazon Web Services has also placed Mechanical Turk under its maintenance category, which means the service will continue operating for current users but will not receive new features.

The move does not immediately shut down Mechanical Turk. Instead, it signals that Amazon is putting the platform into a slower, support-focused phase after years of reduced visibility and growing questions around its role in the modern AI data economy.

A Quiet Shift for a Once Important Platform

Mechanical Turk launched in 2005 as one of Amazon’s earliest experiments in distributed digital labor. The idea was simple: businesses and researchers could post small online tasks, known as Human Intelligence Tasks, and a global pool of workers could complete them for payment.

Those tasks often included data validation, survey responses, image labeling, content moderation, transcription, product categorization, and other work that software could not easily handle at the time. For companies, the appeal was access to a flexible workforce. For workers, the platform offered a way to earn money by completing small jobs online.

In its early years, Mechanical Turk became a key tool for researchers, startups, and companies that needed quick human judgment at scale. It also became widely discussed in debates about online labor, low pay, worker protections, task quality, and the hidden human work behind many digital systems.

What Amazon Has Changed

Amazon’s notice says Mechanical Turk will be closed to new customers from July 30, 2026, while existing users will not be affected by the change. AWS documentation also states that services in maintenance remain available to current customers but are no longer open for onboarding, and AWS does not plan to add new functionality to them.

That distinction matters. Mechanical Turk is not disappearing overnight. Current requesters can continue posting work, and workers can still complete tasks once they are approved by requesters. However, the service will no longer grow through new customer registrations, which could gradually reduce activity on the platform over time.

AWS has said it will continue investing in security and availability improvements, but the absence of new feature development makes the direction clear. Mechanical Turk is being preserved for existing use, not rebuilt for the next phase of cloud, AI, or data-labeling work.

Why Mechanical Turk Mattered

Mechanical Turk arrived before the current wave of AI tools, before large language models became mainstream, and before data labeling became a major industry category. At the time, it helped solve a practical problem: computers were fast at calculation, but weak at many tasks that required simple human perception or judgment.

A person could identify whether an image contained a specific object, decide whether a sentence sounded positive or negative, check whether two records referred to the same product, or answer a survey question. Mechanical Turk turned that type of work into a marketplace.

Over time, the platform became especially important for academic research and machine learning projects. Researchers used it to recruit survey participants, test behavioral questions, and gather labeled data. AI teams used human workers to annotate examples, review outputs, and help build datasets.

That history makes Amazon’s decision notable. Mechanical Turk was not just another side product. It was one of the earliest mainstream examples of human labor being organized through software at internet scale.

AI Changed the Market Around It

The timing is important because the demand for human-labeled data has not disappeared. In many ways, it has grown. AI companies still need human feedback, safety evaluation, preference ranking, red teaming, data cleaning, and domain-specific labeling.

What has changed is the market around that work. Companies now have more specialized options for data annotation, including managed labeling providers, private workforces, expert review teams, and AI-assisted data platforms. AWS itself offers other machine learning and human review services, including SageMaker Ground Truth and Amazon Augmented AI.

Mechanical Turk’s open marketplace model also became harder to position in an era when companies want higher quality controls, stronger privacy safeguards, and more domain-specific annotation. Public crowd work can be useful for simple tasks, but it is less suitable for sensitive data, regulated industries, complex expert review, or high-stakes AI training.

Amazon’s own documentation warns users not to share confidential information, personal information, or protected health information with the Mechanical Turk workforce. That limitation is important because many modern AI workflows involve data that needs tighter governance than a broad public marketplace can provide.

Workers Face Uncertainty

For workers, Amazon’s decision creates a different kind of concern. Existing users are not immediately affected, but the long-term health of a marketplace depends on both sides: requesters who post tasks and workers who complete them.

If new customers can no longer join, the number of new task creators may slowly shrink. That could mean fewer projects, less task variety, and reduced earning opportunities over time, especially if older requesters move to other platforms.

Mechanical Turk workers have already faced years of complaints around inconsistent task availability, low pay, account issues, rejections, and competition from bots or automated tools. The platform’s maintenance status will likely deepen concerns that Amazon is no longer treating MTurk as a major growth product.

Still, the immediate message from Amazon is that current users can continue using the service. Worker payments are also not affected by the change, according to the platform’s help information.

A Slow Retirement, Not a Shutdown

Amazon has not described the move as a full shutdown. That is important for companies and researchers that still rely on Mechanical Turk for ongoing projects. Existing requesters are not being forced off the platform on July 30, and existing workers are not being removed because of the new customer restriction.

But in software terms, maintenance status is usually a clear signal. A product may remain available, but its future becomes limited. Without new onboarding and without feature development, customers often begin planning alternatives.

That is likely what many organizations will do now. Research teams, data-labeling operations, and businesses that still depend on MTurk may need to review their workflows, export key data, check billing arrangements, and compare replacement platforms before the ecosystem becomes thinner.

What Comes Next

The decision reflects a wider shift in the data and AI industry. In the first wave of internet crowdsourcing, the goal was often to access a large number of general workers quickly and cheaply. In the current AI market, the priority has moved toward quality, traceability, privacy, and specialized human feedback.

That does not make human labor less important. It makes it more structured. AI companies still rely heavily on human review, but the work is increasingly handled through managed vendors, expert contractors, internal review teams, and controlled annotation environments.

Mechanical Turk helped prove that distributed human intelligence could be delivered through an online platform. Now the industry it helped shape has moved into a more specialized phase.

The Bigger Picture

Amazon’s decision marks a quiet but meaningful moment in the history of online work. Mechanical Turk was once one of the clearest examples of how the internet could break large human tasks into small digital assignments. It supported researchers, startups, AI teams, and businesses long before today’s data-labeling industry matured.

Its move into maintenance does not erase that legacy. But it does show that the old microtask marketplace model is no longer central to Amazon’s AI and cloud strategy.

For existing users, the service continues for now. For new customers, the window closes on July 30, 2026. For the broader tech industry, the message is larger: the future of human-in-the-loop AI work is not disappearing, but it is moving away from the open crowdsourcing model that Mechanical Turk helped define.