Last Friday evening we noticed an influx of rule-violating comments in /r/AskHistorians. Every 5 minutes or so a user would attempt to answer a question; however, because they were very obviously violating our rules we were able to quickly identify and remove them. This has continued on and off since then. The comments appear to be generated by an AI powered chatbot, like ChatGPT, and the account names follow a generally consistent pattern, so we’re pretty sure we’re dealing with a botnet (although some accounts that seem to follow the posting pattern don’t share the username convention). The accounts are all at least 20 days old, which means we’d need to set an account age filter way higher than we’d like to in order to manage the spam that way. Multiple bots will also post repeatedly on the same post.

So far we’ve been removing, spamming, and banning these accounts and on Friday, reported a large number for ban evasion (which was partially successful in identifying a few as coming from the same source, but not all), but it’s a lot of work and it’s annoying. We were wondering if:

• this was on Reddit’s radar and
• if anyone has developed any successful automod workarounds for this, or similar bots.

We're being light on details out of an abundance of caution, but are also reaching out to mod teams we know are affected, and are happy to share details with subs that think they might be affected too if you shoot us a message via modmail.

edit for grammar