Allow message blocked by automod without adding it to the permitList
A few words can be used for good or bad in many different contexts. When we get a message on a good context and we allow the message, we are adding the word on the permitList which can be dangerous if used on a wrong scenario.
Example: ***.
Good scenario: My sleep schedule is *** right now.
Bad Scenario: This streamer is ******.
If the good message comes first and then the bad message, the second message will not be blocked by automod.
Idea: having a third option, other than accept and deny message, on automod to allow the message on the chat without adding the blocked word into the permitList. Suggestion name: except.

-
Kozaka commented
It is more annoying, that kind of automod (can not be set by streamer!) is on predictions. And my stream language is NOT English. But I want to start a prediction with the text: "Nyerünk vagy sem?" (Do we win or not?). And this automod says the text is inappropriate because of the word "vagy" what means translated "or" from Hungarian and not an abbreviation of the ******. Comeon!
-
Peves commented
My channel is marked as Mature audiences only and i can't make point redemptions that say the F word in them? Why? That doesn't even make sense. It's already marked as a mature channel.
-
Tei2 commented
Recently Twitch had to help with whitelisting "queer", because there were people that used it in a normal way. To represent themselves. They felt comfortable using the word to describe what they are and what they represent. Sadly there are quite a few people that use it in a malicious way. So yes, I agree. The permitList process needs to be looked into.
-
PickledTiger_ commented
Automod is still false flagging perfectly innocent messages. For example, in the chat i moderate we have every filter turned off except for Less Filtering on Race, ethnicity, or religion and Sexuality, ***, or gender and Automod, even after its "update" has flagged things such as "night" when someone was saying "good night" it's flagged "alo" when someone was typing "alone" and it also flagged "H" from someone saying "How". Fix automod to stop flagging the dumbest stuff when all i need is for it to catch people trying to use racial or homophobic slurs.
-
Euther commented
My "Honest" feedback, Like it!
If a streamer has enabled the Mature Content selector, they should continue to have descent customizable automod (AM) selectors, or the choice to disable AM entirely. They (or channel mods) should be able to self moderate. It's their Job!
Also, I would offer adding the ability to filter out (blacklist) specific emotes.
-
RodySV commented
disable the ability to send forbidden words to the chat
-
Kohlrak commented
In order to prevent false positives or to allow someone to stop supporting a channel that over censors, the idea of not informing a user, i would consider, a rights violation. Shadow banning and shadow blocking is one of the most horrid trends we've seen lately.
As an alternative, proper REGEX keys would solve this quickly. I've been testing it with a bot of mine, and so far no false positives, and it accounts for typing something a different way.
-
MrCoolestMan commented
So, like blocked terms be held through automod, even tho there blocked terms, let mods chose either they go through or not, so say someone said the blocked term of "Revenge" and mods dont want it to be used one stream, but mayb accept it another stream.
-
AlmostRobot commented
Currently when someone typing a block term gets the message "Your message wasn't posted due to conflicts with the channel's moderation settings." in some cases this is like "hey, just type it differently" and even if you are use wildcards he can post the message.
Sometimes a message contain a block terms that if that terms wouldn't exist the message would be perfectly normal. Why don't you add the option for some terms to just show like *** and let the message be posted? That's how mods can decide if the message deserve a delete or not, and reduce the chance for users that type the block terms alternate.