Skip to main content
Cloud / AWS / Products / Amazon SageMaker Ground Truth: Label Training Data

Amazon SageMaker Ground Truth: Label Training Data

Amazon SageMaker Ground Truth builds labeled training datasets using human workforces. Closed to new customers since July 2026.

Machine Learning
This service is deprecated or retired

Per the documentation, Amazon SageMaker Ground Truth is no longer open to new customers; the AWS service availability announcement from June 2026 names July 30, 2026 as the end of new customer access. Existing customers can continue to use the service; per the documentation AWS continues to invest in security and availability but does not plan to introduce new features. Per the same announcement, Amazon SageMaker AI Ground Truth Plus reached end of support on June 30, 2026.

Pricing Model Rates per provider (see official pricing information)
Availability Regional availability per provider (see official documentation)
Data Sovereignty Region and sovereignty details per provider
Reliability SLA per provider (see official SLA page) SLA

What is Amazon SageMaker Ground Truth?

Amazon SageMaker Ground Truth is a capability of Amazon SageMaker AI for building labeled training datasets for machine learning models. You combine human workers with machine learning; the output can be used to train your own models or as a training dataset for an Amazon SageMaker AI model.

Important lifecycle note: Per the documentation, Ground Truth is no longer open to new customers. Existing customers can continue to use it, and no new features are planned. The AWS service availability announcement from June 2026 names July 30, 2026 as the end of new customer access. New initiatives should therefore plan for a different solution.

Core Features

  • Workforce of your choice: Amazon Mechanical Turk, a private workforce of your employees or contractors, or a vendor company from the AWS Marketplace
  • Built-in task types: Pre-built task types for common labeling tasks
  • Custom labeling workflows: Your own UI and tools for workers, built with HTML 2.0 components
  • Labeling UI templates: The SageMaker AI console provides built-in templates to get started
  • Automated data labeling: Machine learning decides which data needs human labeling
  • Streaming labeling jobs: Perpetually running jobs that hand new dataset objects to workers in real time
  • Data in Amazon S3: The data to be labeled, the input manifest, and the output manifest live in the bucket
  • Logging: Labeling job events appear in Amazon CloudWatch under /aws/sagemaker/LabelingJobs

Typical Use Cases

Ongoing projects of existing customers: Keep existing labeling pipelines running while the service remains available to existing customers.

Bounding boxes and classification: Annotate image data using built-in task types.

Sensitive data in house: A private workforce of your own employees labels data that must not leave the organization.

Continuous labeling: Streaming labeling jobs for data streams that continuously deliver new objects.

Benefits

  • Choice between public, private, and vendor workforces
  • Built-in task types shorten labeling job setup
  • Custom interfaces possible for domain-specific tasks
  • Automated data labeling can reduce manual effort
  • Integration with Amazon S3 and Amazon CloudWatch

Integration with innFactory

As an AWS Reseller, innFactory supports you around Amazon SageMaker Ground Truth: keeping existing customers’ labeling pipelines running and optimized, evaluating alternatives for new initiatives given the closure to new customers, and building training data and MLOps processes on AWS.

Typical Use Cases

Ongoing labeling projects of existing customers
Building labeled training datasets
Custom labeling workflows with a private workforce
Streaming labeling jobs

Technical Specifications

Automation Automated data labeling uses machine learning to decide which data needs to be labeled by humans
Data Datasets are stored in Amazon S3 buckets with an input and an output manifest file
Interfaces Labeling UI templates; the SageMaker AI console provides built-in templates, and custom ones are built with HTML 2.0 components
Logging Labeling job events appear in Amazon CloudWatch under the /aws/sagemaker/LabelingJobs group
Positioning A capability of Amazon SageMaker AI; the output can be used as a training dataset for an Amazon SageMaker AI model
Task types Built-in task types as well as custom labeling workflows with your own UI
Workforces Amazon Mechanical Turk, a private workforce of your employees or contractors, and vendor companies from the AWS Marketplace

Frequently Asked Questions

Is Amazon SageMaker Ground Truth still available?

Per the documentation, Amazon SageMaker Ground Truth is no longer open to new customers. Existing customers can continue to use the service as normal. Per the documentation, AWS continues to invest in security and availability improvements but does not plan to introduce new features. The AWS service availability announcement from June 2026 names July 30, 2026 as the end of new customer access.

What does Ground Truth do?

Per the documentation, Ground Truth helps you build high-quality training datasets for machine learning models. You combine human workers with machine learning to create a labeled dataset, which you can use to train your own models or as a training dataset for an Amazon SageMaker AI model.

Which workforces can be used?

Per the documentation, you can use the Amazon Mechanical Turk workforce of over 500,000 independent contractors worldwide, a private workforce of your employees or contractors, or a vendor company from the AWS Marketplace that specializes in data labeling. Per the June 2026 announcement, Amazon Mechanical Turk itself is closed to new customers as of July 30, 2026.

What is automated data labeling?

Per the documentation, automated data labeling is an optional Ground Truth process that uses machine learning to decide which data needs to be labeled by humans. It may reduce labeling time and manual effort.

What are streaming labeling jobs?

Per the documentation, a streaming labeling job sends new dataset objects to workers in real time. The job runs perpetually, and workers continuously receive new objects to label as long as the job is active and new objects are being sent to it.

What about Ground Truth Plus?

Per the AWS service availability announcement from June 2026, Amazon SageMaker AI Ground Truth Plus reached end of support on June 30, 2026.

Note: All product information on this page has been compiled with care, but is provided without guarantee and may be outdated or incomplete. Cloud services evolve rapidly — features, pricing, SLAs, and availability change frequently. Authoritative and up-to-date information can only be found on the official product page of AWS (official documentation). This page does not represent an offer by AWS.

AWS Cloud Expertise

innFactory is an AWS Reseller with certified cloud architects. We provide consulting, implementation, and managed services for AWS.

Ready to start with Amazon SageMaker Ground Truth: Label Training Data?

Our certified AWS experts help you with architecture, integration, and optimization.

Schedule Consultation