As a Roblox user, it is currently hard to trust the usage of the “Chat without filters” feature for trusted connections to deeply express my feelings on chat while fearing the risk of Roblox’s automated filter
Recently, there has been a ongoing ban-wave for the usage of the unfiltered chat for friend/party chats that targets words Roblox does not tag according to their community guidelines, terms like “Discord” are sometimes tagged or untagged which had led to account suspensions ranging from “Giving PII” to Literal Child Endangerment
The main thing I would want about this suggestion is to at-least add proper filtering for slurs, heavy swear words, NSFW and off-topic platforms to prevent hundreds to thousands of users from being suspended for a action they have now regretted, standard swearing like the F word could be allowed if used as a informal expression and not direct harassment. So TLDR add filtering for words that would violate Roblox’s Community Standards
If this suggestion were be approved and came out, It should at-least give users who previously said bad things in archived and trusted connections chats a second chance and a warning on using those specific words via filtering previous words (NSFW, slurs all of those that the AI would deem unsafe) that would violate Roblox’s community standards in every friends/party chat involving them
(unless of course if it involves talk that could escalate into Law Enforcement being called)
I believe the purpose of the feature is intended to make the filter less strict to lower the possibility of rule-abiding messages being blocked. Messages still need to follow the Community Standards though which means swearing is disallowed.
Yes I agree, however they had poorly designed it and was clearly a rushed update (by that, they do not tag any slurs, extreme swear words etc) without adding a proper boundaries on what can be said or not
They market it as “more natural conversations” but when you try to act naturally you could be unknowingly break their long community guidelines, you could be facing account punishments, I’m saying there should at-least be a proper barrier of what can be said and what can’t
I personally think that the Community Standards are pretty understandable. They do also state that you have to still comply with their Community Standards when using it.
Even though users have to follow their Community Standards,
They’re practically expecting absolute perfection from the user to not break any of their rules without proper filters or safety nets in the “Chat without filters” which essentially sets a trap for people who might accidentally type a bad word without knowledge
It honestly varies whether the AI is feeling generous or not, I’ve had a friend who had gotten a 14 day ban for saying the F word once with no prior moderation history
I’m fairly certain Moderation also may account for other things besides prior moderation logs and the content. I may be wrong though. I do understand what this post is saying though.
Deleting my bumps as i feel like they’re useless, i’m going to provide more information as i discover issues to back this suggestion up, first up from what i’ve been told,
this is how the ban message would look like if you had gotten banned for a party chat violation, don’t want to gloat about this too much, if the suggestion were to be implemented, it could reduce frustrating false-positive bans while balancing user safety to any engineers/staff reading this
