9

So this is a follow up to this question I asked when I thought the spam flood was just a one-night ordeal (which was very naive, it turns out): Is there any way to stop the flood of spam coming in right now?.

Most of these spam questions have "credi" and/or "loan app" and/or a few other indicator words/phrases, with varying degrees of capitalization, spacing, and punctuation, in their title after being edited. Here's a screenshot of my browser history to demonstrate, though I'm sure to members here this is probably unnecessary (yes, I have forty-something "helpful flags" here, all from the last few weeks): enter image description here.

So I'm thinking, is there a way to automatically flag questions with "credi", "hisab", "loan app", or "helpline" in their title and that are written and edited by new contributors with 1 rep (all I've seen have been by "new contributor" 1-rep users) for immediate moderator attention when they're edited before releasing them back to the main site? (They seem to all contain copies of pieces of legitimate questions, and the titles afterward are edited to contain the spam.) I get that legitimate questions that match this description are maybe remotely possible, but moderators could simply release these to the main site and delete the others.

So, my idea:

Immediately send questions for moderator attention and remove from main site that:

  • are edited to contain "credi", etc. in the title,
  • are posted by "new-contributor" users with 1 rep

Have mods review them and, if legitimate (unlikely but conceivably possible), release to main site, otherwise delete them.

Is this possible? If so, advisable?

8
  • 5
    I'd also add - for the folks from the company looking at this - this is a relatively new user to the network. Is a page full of spam what we want a new user to see? Commented Aug 3, 2025 at 4:25
  • 2
    There are times at night where there are dozens of spam that exist at once. Granted they are deleted relatively quickly, but it can take at least 5 minutes to spam flag this content, because if the rate limit on flags. Commented Aug 3, 2025 at 12:04
  • "The behaviour of my PC is incredible". Please don't block that person as a spammer 😀😀😀 Commented Aug 4, 2025 at 7:38
  • @Dominique, that's my point - those questions would be released into the wild if they're legit. Commented Aug 4, 2025 at 18:31
  • Maybe force moderator approval for anything with lots of hindi characters or phone-numbers-looking chars. Commented Aug 8, 2025 at 1:52
  • I think the hellban might actually be needed. Commented Aug 8, 2025 at 1:54
  • I would've thought that the SE spam filter would already have such a functionality. Commented Aug 10, 2025 at 1:07
  • Marking this deferred same as this request. Will review again once the upcoming anti-spam tooling has been live for a while, to determine if a change like this is still needed. (Given the nature of spam though, I suspect this would be quickly evaded even if implemented.) Commented Oct 2, 2025 at 21:06

2 Answers 2

Reset to default
15

Funny thing. We've a filtering system and asked about it and this would handle most of the non gibberish spam. Blocking Hindi and certain keyboards (in a non case sensitive way) might help a bit . We've been told they can work around it.

I've handled a large chunk of the issue since it happens during my timezone, and honestly outside a massive bump in per account flags, support feels lacking.

Honestly between this, and certain other current events going on. I'm a little burnt out, so even if the feature request as posted doesn't happen, dear reader, please take this as a call for help. While we've been dealing with this effectively, at some point, if there's a massive spam wave, or we lose folks, we might not be able to keep up.

5
  • 3
    "Blocking Hindi and certain keyboards (in a non case sensitive way) might help a bit . We've been told they can work around it." well, unfortunately yes. I'm pretty sure they'd just adapt the spam messages to not hit the filters :/ Commented Aug 3, 2025 at 8:11
  • 4
    That's the argument I get but the full language block was useful for chinese spammers. While we can't get 100% coverage - we're dealing with hundreds of spammers with multiple posts getting in. Yeah, we'd need to adapt the filters periodically but considering the sheer amount of spam and complaints about it, it might be nice to at least try it again. Commented Aug 3, 2025 at 8:20
  • 4
    Something has to be done. Anything, even if they change their tactics, that would be a win to the hundreds of spam questions, answers, and comments being submitted daily. There was so much in the last few days. I just didn’t even bother flagging. Commented Aug 3, 2025 at 12:06
  • 2
    "I didn't bother flagging" <-- that right there is an indicator that SE is at risk of losing a valuable asset. Commented Aug 3, 2025 at 19:32
  • 1
    I flag for a while then get tired of it and move on to something else. User flagging works but is far from ideal. Commented Aug 4, 2025 at 2:33
2

Just a question: how will we discuss this subject?

Imagine a tool is created, removing all questions who obey to the conditions we have set here. At one side, this might cause obsolete questions to be removed.

On the other hand, the discussion also explains to spammers how not to create their posts, and by using a very slightly different format/wording/..., they might actually know how to write spam, based on our discussion.

How might we manage agreeing on a good spam filter without informing spammers about it?

1
  • 5
    If we get agreement to do it - it would be trivial to get a private space - chat or even a team to discuss it. Practically, we could have staff and mods co-odinate, and we've a reasonable corpus of data from metasmoke or community reports - as with many problems, the first step is accepting its a problem and that we need resources to tackle it . Commented Aug 7, 2025 at 15:02

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.