AI Guardrails is an Alibaba Cloud service that intelligently moderates multimedia content such as images, videos, and text in various scenarios. This service helps you effectively reduce the risk of content violations. AI Guardrails is based on AI algorithms and cloud computing technologies to identify and label inappropriate or illegal multimedia content. AI Guardrails supports the moderation of objects such as images and videos in various industries and business scenarios. It can help you improve content moderation efficiency, quality of platform content, and user experience.
Content Moderation is a pioneer in the field of content moderation. It originates from Alibaba Cloud's many years of security technology accumulation and relies on the management and control experience of Alibaba Cloud, Taobao, Alipay, and other platforms to provide enterprises with mature and lightweight content moderation solutions.
Feature description
Alibaba Cloud AI Guardrails is a pioneer in the field of content moderation. It originates from Alibaba Cloud's many years of security technology accumulation and relies on the management and control experience of Alibaba Cloud, Taobao, Alipay, and other platforms to provide enterprises with mature and lightweight AI Guardrails solutions. AI Guardrails helps enterprises and developers quickly detect various risks in text, images, and videos in complex and ever-changing Internet environments.
Currently, the AI Guardrails API is available.
The AI Guardrails API moderates text, images, and videos for pornography, politically sensitive content, terrorism, ads, and junk information. Using systematic methods, the API offers powerful capabilities such as content moderation, labeling, and custom configuration to achieve desired effects and meet your tailored requirements. To use the API, you must integrate it into your applications.
The applicable industries include but are not limited to all user-generated content (UGC) platforms such as video websites, live broadcast platforms, social platforms, media platforms, vertical communities, forums, e-commerce websites, storage platforms, and content delivery network (CDN) platforms, and other platforms that require moderation for website content.
The problem with manual moderation
Platforms that depend entirely on human moderators struggle to keep up with the volume and speed of user-generated content (UGC). Manual-only pipelines are slow, expensive to scale, and inconsistent. Unenforced content policies damage user experience and brand reputation.
Content Moderation addresses this by running AI-based checks across all submitted content. Human moderators can then focus on a smaller, higher-priority subset of flagged content—rather than reviewing everything—significantly reducing the volume of content that requires human attention.
How it works
Integrate the Content Moderation API into your application. The API accepts images, videos, and text, runs automated moderation checks, and returns labels for each violation category. Act on the results by approving content, rejecting it, or escalating it for human review.
Key capabilities
| Capability | Description |
|---|---|
| Image moderation | Detects pornography, politically sensitive content, terrorism, ads, and junk information in images |
| Video moderation | Applies the same detection categories to video content |
| Text moderation | Identifies pornography, politically sensitive content, terrorism, ads, and junk information in text |
| Labeling | Returns labels for each detected violation |
| Custom configuration | Lets you adjust moderation thresholds and rules to match your platform's policies |
Use cases
Content Moderation is designed for platforms that handle user-generated content (UGC), including:
Video websites and live streaming platforms — screen uploaded videos and live streams for violations before or as they publish
Social platforms and forums — moderate user posts, comments, and messages in real time
E-commerce websites — review product listings and seller-submitted images for policy violations
Media platforms and vertical communities — enforce editorial standards on text and visual content
Storage platforms and content delivery network (CDN) platforms — scan stored assets for compliance before delivery
Get started
Quickstart: Get started with the Content Moderation API
Concepts: Learn how the detection categories work
How-to guides: Configure custom moderation rules and thresholds for your platform