Trust Center
Local, Hybrid & Cloud Processing
Chameleon uses different processing modes for different workflows. Privacy claims are mode-specific rather than platform-wide.
Chameleon AI does not automatically upload a user's full private archive, document library, or local memory to cloud AI infrastructure. What is processed depends on the selected workflow and processing mode.
Processing Mode Definitions
The public wording below is the intended privacy boundary for each mode.
Offline Local Mode
Implemented- Processing location
- Supported local workflow on the user's device / local environment
- Cloud AI provider
- No cloud AI provider is contacted for the supported local workflow
- What is sent
- No payload is sent to a cloud AI provider for that supported local workflow
- Approval
- Cloud approval is not required because the workflow remains local
Hybrid Mode
Implemented- Processing location
- Local first + approved cloud infrastructure
- Cloud AI provider
- Only after user approval
- What is sent
- Only user-approved content is sent to approved cloud AI infrastructure
- Approval
- User approval is required for the cloud step
Cloud Mode
Implemented- Processing location
- Chameleon server-side routing + approved AI infrastructure
- Cloud AI provider
- Yes
- What is sent
- The approved request payload is processed through server-side provider routing
- Approval
- Controlled by the applicable workspace, product, and request flow
Offline Local Mode
- +Designed for supported workflows that can be executed locally
- +No cloud AI provider is contacted for that supported local workflow
- +The local workflow does not require provider-side AI processing
- +Use depends on the installed local capability and the specific workflow
Hybrid Mode
- ~Starts from a local or application-controlled workflow
- ~Cloud processing occurs only after user approval
- ~Only the content approved for that cloud step is sent
- ~The full private archive is not uploaded automatically
Cloud Mode
- ~The approved request payload is sent through Chameleon's server-side routing
- ~Provider credentials remain server-side and are not exposed to the browser
- ~The exact payload depends on the product workflow and user action
- ~The full private archive is not uploaded automatically
Dedicated private-cloud or zero-data-retention arrangements are not represented here as generally available product modes. Where required, they are subject to separate provider, contract, and implementation review.