Privacy & security

Understand what is sent to AI providers

Review what LobeWork sends, what remains local, and how to make a deliberate choice before using a connected model.

8 min readBeginnerUpdated 2026-07-26For LobeWork 1.0+
01

LobeWork organizes the request

Your project may contain more information than a provider needs. Build a focused context package for the current task.

Build a focused context package for the current task.

02

Providers receive what you submit

Text, files, images, prompts, and tool results included in a provider request are processed under that provider’s terms and settings.

03

Local data can remain local

Files and notes that are not attached to a provider request remain part of the local workspace.

04

Review sensitive content

Remove credentials, private personal data, confidential customer information, and unrelated material before sending.

05

Use local or offline workflows where appropriate

A local model can reduce cloud processing for supported tasks, but it may have different capabilities and hardware requirements.

NEED A GUIDED ANSWER?

Ask LobeWork about this guide.

The support assistant uses the same published guide library and will link you back to the most relevant article.

WAS THIS GUIDE HELPFUL?Help us improve the documentation.