Google Cloud Model Armor
Managed screening for prompts and model responses, including potential prompt injection and sensitive data exposure.
Published by Google Cloud
How to use it
Define inspection policies for AI traffic and apply them through supported Google Cloud integrations or the service API. It is intended for teams that want a managed inspection layer around one or more AI applications.
Decide how an application should respond to a flagged interaction and where sensitive data is allowed to go. Google publishes pay-as-you-go and subscription options.
- Access
- Paid cloud service