Good evening!
Recently, bots spamming Discord have been infiltrating on a regular basis. Upon investigating their behavior, although I haven’t reached a full conclusion yet, it seems like these bots are using AI for their spamming activities.
The content of their messages promotes a certain NFT (details unknown) and a service called RENZSWAP. Since they are persistently spamming using bots, both of them are scams, so please be careful. These bots proliferate accounts to spam Discord, and given the situation, there’s a high chance you’ll encounter such messages while browsing through Discord.
Their methods are becoming increasingly sophisticated, and judging from their behavior, it seems to be AI-driven. Even using regular expressions to block certain words to avoid spam is ineffective… because the AI can instantly predict the banned word candidates and generate new sentences with the same nuances while avoiding those words. It seems that we’ve entered an era where these old methods are no longer effective. Therefore, the security features Discord has added for word filtering are no longer sufficient.
In the case of RENZSWAP, for example, the bot repeatedly rewrote the message into what seemed like an AI-generated sentence and posted persistently. Since AI predictions are instantaneous, the accounts proliferate rapidly (perhaps it learned when to recreate them), surpassing the speed at which bans are applied. I’m currently observing how Discord will respond to this.