Ring AI Video Descriptions and Single Event Alert (Beta)
Ring AI Video Descriptions and AI Single Event Alert help you understand what's happening around your home. Learn how AI-powered features can enhance your Ring camera and doorbell experience.
Requirements and availability
To use Ring AI Video Descriptions and Single Event Alert, make sure you meet the requirements:1
- Only available to mobile devices with the primary language set to English
- Not available for Ring devices in Illinois due to state legislation
- Not compatible with Ring Edge, Ring Alarm, Ring Alarm Pro (non-motion footage), or Ring Car Cam
Important note: Beta features are new or experimental. While you get early access to innovative features, results might vary as we continue making improvements based on user feedback.
Managing AI Video Descriptions and AI Single Event Alert
How AI Video Descriptions work
AI Video Descriptions adds detailed text to your Video Preview Alerts. When your Ring device detects motion or someone rings your doorbell, you'll get a notification with both an animated preview and an AI-generated description of the beginning of the activity. When AI Video Descriptions is on, you'll see text descriptions in your Event History. These descriptions help you quickly scan through your recordings. Turning off the feature will remove all descriptions. You can activate or deactivate AI Video Descriptions for individual devices at any time. You will need to activate the AI Video Description feature on any new Ring devices that you add to your account.
To activate or deactivate AI Video Descriptions:
Open the Ring app to the main dashboard.
Tap more (•••) on the camera you want to access.
- TapAlert Settings.
- TapVideo Descriptions.
Understanding description accuracy
AI Video Descriptions might occasionally be inaccurate. Common reasons include:
- Poor lighting or limited visibility
- Fast-moving objects or people
- Complex or overlapping activities in the scene
- Weather conditions affecting video quality
- Objects or people too far from the camera
How AI Single Event Alert works
AI Single Event Alert reduces notification clutter by combining similar alerts from the same camera within a set time period. You can turn AI Single Event Alert on or off in the Ring app. If you already had AI Video Descriptions activated, AI Single Event Alert will be auto-enabled on any new devices added to the account.
To turn AI Single Event Alert on or off:
Open the Ring app to the main dashboard.
Tap more (•••) on the device you want to access.
- TapAlert Settings.
- TapAI Single Event Alerts.
Important note: Expect a brief delay in notifications when using AI Single Event Alert. The system needs extra time to check for similar events, but this helps reduce notification clutter. When your Ring device detects motion or someone rings your doorbell, you'll get a notification with both an animated preview and an AI-generated description of the beginning of the activity.
How the features work together
AI Single Event Alert relies on the content from Video Descriptions to function. If you turn off Video Descriptions, AI Single Event Alert automatically turns off.
Finding your recordings
With these AI features on, all recordings are saved in your Event History. Similar events that happen around the same timeframe are grouped together in a carousel format for easy viewing. You can use the Event History Filter to view them without grouping. Learn more about viewing your Event History. Single Event Alert automatically groups similar videos that happen around the same timeframe in your Event History to help you find footage faster. This grouping feature only works when Single Event Alert is turned on. Videos are arranged chronologically, with the most recent event appearing first in each group.
We are committed to Responsible, Trusted AI
We are committed to ensuring we are building safe and responsible AI-powered products and features throughout our design, development, deployment, and operation. To help ensure Video Descriptions generates reliable and appropriate content, we built a protection system. The system consists of Vision-Language Model (VLM) built-in and real-time guardrails to ensure appropriate descriptions are provided to customers.