Introduction to Content Guard AI ✨

Content Guard is your automated content watchdog. Using advanced image analysis and Artificial Intelligence, it constantly monitors screenshots from your devices and automatically classifies them into categories. If it detects an error or prohibited content, it notifies you instantly.

Core concepts

Before you start monitoring, it is important to understand three basic terms:

1. Category: A label that defines the status of the content (e.g., "Menu Board" or "Error"). Categories are classified as either Valid (all good) or Invalid (issue detected).
2. Content Guard Item: A specific rule or reference image used by the engine to recognize content.
3. Alert Rule: The logic that triggers a notification when a category appears or disappears.

Detection methods: Image vs. Prompt

When creating your own rules, you can choose between two primary technologies:

1. Image-Based detection

This method looks for visual similarity (a digital fingerprint).

• When to use: For static content that always looks the same—e.g., a specific menu board, a logo, or a frozen system screen.
• Benefit: Extremely fast and highly accurate for identifying identical or near-duplicate content.

2. AI-Powered detection (Prompt-based) ✨

Instead of an image, you define a natural language description (a Prompt). The AI "looks" at the screenshot and understands its context.

• When to use: For dynamic content where elements change positions, or when you want to monitor general concepts (e.g., "There should be a burger visible" or "I don't want to XXXs logo appears").

AI Helper✨

If you are not sure how to write the perfect AI instruction, upload a sample photo to the AI Helper. It analyzes the image and automatically generates a professional technical description (Positive or Negative prompt) for you.

System-managed content guard items and categories

To simplify your start, Content Guard includes pre-defined categories and items that are ready to use out of the box. These are "Read-Only" templates managed by the system.

What is already prepared for you?

• Blank screen (Invalid): Detects black, white, or empty screens caused by signal loss or rendering failure.
• No applet (Invalid): Detects states where the system is running, but the specific application content is missing.
• Proprietary timer off (Invalid): Detects system overlays or timers that indicate playback has stopped.

You can select these managed categories immediately when creating your Alert Rules.

How to set up Alert Rules

Connecting Content Guard to the Alert Manager is the key. In the Alert Rules section, create a new rule of type Content Guard:

1. Select category: Choose one specific category (either your own or a system-managed one).
2. Set frequency:
   • „Not appear“: If your category (e.g., "Live Promo") does not appear at least once during 60 minutes, the system triggers an Alert. (Best for ensuring required content is running).
   • „Appear“: If the category (e.g., "Blank Screen") appears 1x (or more) during 60 minutes, the system triggers an Alert. (Best for instant error detection).

Monitoring & Tag management

In the Monitoring Dashboard, you will see your Content Guard categories and their validity (green/red tags) directly on the screenshot thumbnails.

In the Tag Management section, you will also find a Content Guard Usage counter for every tag. This shows you exactly how many detection items (Hashes or Prompts) are currently using that specific tag, helping you avoid accidental deletions of active monitoring rules.

Supported Platforms

AI ContentGuard works across all major platforms supported by signageOS:

Platform	Version
Android	4.14.6
Tizen	2.11.1
WebOS	2.10.2
Brightsign	2.4.1
Linux	2.7.0

Quick Tip: Check out our interactive Onboarding Tour directly in the app (click the question mark How does it work?). It will walk you through creating your first AI Prompt step-by-step!