[GH-ISSUE #7475] [FR] Make it possible to disable AI writing functionalities using toggle to prevent accidental activation #3331

New issue

Closed

opened 2026-03-23 21:29:23 +00:00 by mirror · 1 comment

mirror commented

2026-03-23 21:29:23 +00:00

Owner

Originally created by @rampa3 on GitHub (Mar 6, 2025).
Original GitHub issue: https://github.com/AppFlowy-IO/AppFlowy/issues/7475

Description

I would like to request possibility to disable AI writing features of AppFlowy. Reason is that it is too simple to activate them, and make the AI spit a ton of unwanted text in manually typed document on misclick. This either clutters the currently edited text, or even overwrites it when such misactivation happens.

Impact

Users that don't want to use AI writing features would prevent accidental activation of them, and possible ruining of their documents this way. Could also be useful for those who keep sensitive data in their notes, and don't want to leak those to the AI providers, while also not wanting to run local inference API.

Additional Context

One misclick made the AI writer overwrite a big part of my university notes I wrote into AppFlowy with a long generated response. I am not interested in the AI writing features in this one workspace, and with how simple it is to trigger them, there is high possibility of ruining the currently edited document if user misclicks.

Originally created by @rampa3 on GitHub (Mar 6, 2025). Original GitHub issue: https://github.com/AppFlowy-IO/AppFlowy/issues/7475 ### Description I would like to request possibility to disable AI writing features of AppFlowy. Reason is that it is too simple to activate them, and make the AI spit a ton of unwanted text in manually typed document on misclick. This either clutters the currently edited text, or even overwrites it when such misactivation happens. ### Impact Users that don't want to use AI writing features would prevent accidental activation of them, and possible ruining of their documents this way. Could also be useful for those who keep sensitive data in their notes, and don't want to leak those to the AI providers, while also not wanting to run local inference API. ### Additional Context One misclick made the AI writer overwrite a big part of my university notes I wrote into AppFlowy with a long generated response. I am not interested in the AI writing features in this one workspace, and with how simple it is to trigger them, there is high possibility of ruining the currently edited document if user misclicks.

mirror closed this issue

2026-03-23 21:29:23 +00:00

mirror commented

2026-03-23 21:29:23 +00:00

Author

Owner

@rampa3 commented on GitHub (Apr 4, 2025):

This issue was created in a moment of anger when my notes got overwritten by AI after a misclick, and I don't think it is useful for others.

@rampa3 commented on GitHub (Apr 4, 2025): This issue was created in a moment of anger when my notes got overwritten by AI after a misclick, and I don't think it is useful for others.