Reporting suspicious new user bots

[J.Quondam]

Often when a bot joins, it has a telltale sign that violates TOS (eg, inappropriate links, logos in profile/avatar, hidden links, etc), so it's easy to flag.
Yesterday, though, i noticed one join that I'm 99% certain is a bot (gut feeling, based on likely AI gen text), but it hasn't done anything wrong afaict, so I didn't report.

Do mods want to get reports for suspicious new users like that, or do you prefer to wait until one explicitly breaks the rules before getting a report?

 

[Umbran]

Often when a bot joins, it has a telltale sign that violates TOS (eg, inappropriate links, logos in profile/avatar, hidden links, etc), so it's easy to flag.
Yesterday, though, i noticed one join that I'm 99% certain is a bot (gut feeling, based on likely AI gen text), but it hasn't done anything wrong afaict, so I didn't report.

Do mods want to get reports for suspicious new users like that, or do you prefer to wait until one explicitly breaks the rules before getting a report?

We were just talking about that among ourselves.

I think that if we start banning users over "gut feelings", there will be a ton of false positives.

I expect that the battle against bots lies not in our banning an occasional new poster for stilted language in slightly off-topic posts, but in policies for account creation or posting links.

 

[Whizbang Dustyboots]

AI bots have gotten a lot better very suddenly, across all platforms. Someone has clearly rolled out some new products.

I think responding due to problematic behavior is probably the right way to go. Lord knows we have plenty of perfectly human posters who operate off of limited scripts as well, and we don't ban them.

 

[J.Quondam]

Will do. I'll stick to reporting based on visible behaviors rather than suspicious language.
Thanks for the quick reply.

 

[Wofano Wotanto]

I saw a frighteningly effective one over on a company forum the other day. It was good enough that only the inclusion of dubious links in their two posts really made me doubt it was just someone with ESL grammar issues, and even then I wasn't sure - they weren't obvious scam sites although I didn't follow them up for safety's sake. I noted the mods removed the account and posts this morning so I guess they were more confident about it being a bot.

Some scary stuff out there already, and it'll only get worse.

 

[CleverNickName]

We were just talking about that among ourselves.

I think that if we start banning users over "gut feelings", there will be a ton of false positives.

I expect that the battle against bots lies not in our banning an occasional new poster for stilted language in slightly off-topic posts, but in policies for account creation or posting links.

There are a few things we can do as users, too. Here are a few basic online safety tools that everyone with a corporate job already knows; feel free to add more:

1. Be wary of new accounts. Lately, I assume that any account that is less than a week old is actually a bot.
2. Always check a link before clicking on it. Reading the link's mouse-over text is good, right-clicking and viewing the link directly is best.
3. Don't click on social media links (Twitter, TikTok, YouTube). Like, ever.
4. If you absolutely must click the link, copy the link to your clipboard first, then paste it into the browser line. Look for trackers, redirects, and other nonsense, and delete the link if you see them. (be especially vigilant about colons, pound-signs, percent signs, and multiple websites in a hyperlink.)
5. Learn how to use a VPN, and configure it properly.
6. Use Ghostery (or other HTTP/tracker blocker)

 

[Morrus]

So this is going to be a thing that happens more and more, and I don’t know how that will look in the long term. The bots will get better and better and appearing natural. Right now some platforms are deluged with bots, and we don’t want that to happen here. But it’s going to get to the point where we just can’t tell.

I guess on the other side of the coin, we’re not the only platform which will have to deal with this problem and I suspect the big dogs in social media are working on ways to combat this.

 

[Whizbang Dustyboots]