When implementing AI into our filters, we have either an internal or an external integration. The internal integrations leverage the resources built into a computer, such as the graphics processing unit (GPU), the tensor processing unit (TPU), or the Neural Processing Unit (NPU). In these cases, the processing happens entirely within the computer running the filter, and the filter does not transmit any information to outside services.
For external integrations, however, we leverage the service hosted by stability.ai. In those cases, information from your project, including images and text prompts may be transmitted to remote servers in order to handle the processing. When that happens, the remote servers may keep a history of the requests, including the images and any associated prompts and responses. For details, please see the Stability AI Terms of Service.
Please note that unless you explicitly enroll in these services and complete a separate onboarding process–which includes agreeing to a separate terms of service–you will not have access to these features from within our software.