What Data Might Be Processed by Service Providers?
Understand data flow and public disclosures based on the features you use.
Processing Scope Varies with Features
When using Hackcat, messages, files, related conversation context, and tool results may be used to generate responses; account, device, usage, and diagnostic information may also be processed for authentication and operational needs. When using search, web reading, or tool execution, data related to those features may also be involved. The scope should be understood based on the actual features used, rather than assuming only the final question is processed.
Refer to Public Disclosure Pages
The Privacy Policy explains data collection, usage, disclosure, and retention; the Security and Trust page supplements the services relied upon and categories of processing. When reading, first identify the features you care about, then verify what data they involve. Terms may differ among providers, so do not apply one service’s practices to all features.
- Confirm which features you are using this time, such as chat, files, search, or execution.
- Review the Privacy Policy and Security and Trust page.
- Narrow your input scope according to the sensitivity of your materials.
Maintain Accuracy on Unconfirmed Matters
Do not equate local directories, local tools, or a certain login method with all data staying on the device. Also, do not assume all model providers make the same commitments regarding training and retention. When handling sensitive materials, first verify the latest public disclosures, then decide whether to upload, excerpt, or anonymize; details that cannot be confirmed should be kept as pending items.