Brave currently has a severe problem with interpreting Windows 11 group policy when it comes to Leo AI.
On our systems we encountered the following behaviour:
We set “Disable AI Chat” in Classic Administrative Templates (ADM)/Brave Software/Brave/Brave Software settings to “Enabled”. The Description says :
If this policy is set to true, AI Chat will always be disabled.
Now checking brave://policy/ tells us, Brave does indeed see this policy. It’s just configured the wrong way. Inside brave the setting is not disabled, the flags name is BraveAIChatEnabled.
This means, setting the policy to true/enabled does enforce the AI being enabled, despite the description telling otherwise. Setting the group policy to disabled does then disable the AI chat.
Please fix this issue soon, this behaviour violating group policies is a severe security issue.
Indeed, the value is named “Disable AI Chat” in group policy, but Enable in Brave documentation and settings.
This can easily cause confusion. So you should either rename the group policy setting or interpret it correctly. I attached a screenshot to clarify how this setting is displayed in Windows’ group policy editor.
As you can see in the screenshot, you have sometimes “Enable” and sometimes “Disable”, which I generally regard as “bad habit” because why would you confuse users by making one switch enable and the other one disable settings?
Edit:
Here’s brave://policy to clarify even more. There its Enabled=false; that’s the result with Disable=false in group policy:
