Over the past 5-6 months, I've been noticing a lot of new accounts spinning up that look like this format:
https://instance.xyz/u/gmbpjtmthttps://instance.xyz/u/tjrwwiifhttps://instance.xyz/u/xzowaikv
What are they doing?
They're boosting and/or downvoting mostly, if not exclusively, US news and politics posts/comments to fit their agenda.
Edit: Could also be manipulating other regional news/politics, but my instance is regional and doesn't subscribe to those which limits my visibility into the overall manipulation patterns.
What do these have in common?
- Most are on instances that have signups without applications (I'm guessing the few that are on instances with applications may be from before those were enabled since those are several months old, but just a guess; they could have easily just applied and been approved.)
- Most are random 8-character usernames (occasionally 7 or 9 characters)
- Most have a common set of users they're upvoting and/or downvoting consistently
- No posts/comments
- No avatar or bio (that's pretty common in general, but combine it with the other common attributes)
- Update: Have had several anonymous reports (thanks!) that these users are registering with an
@sharklasers.comemail address which is a throwaway email service.
What can you, as an instance admin, do?
Keep an eye on new registrations to your instance. If you see any that fit this pattern, pick a few (and a few off this list) and see if they're voting along the same lines. You can also look in the login_token table to see if there is IP address overlap with other users on your instance and/or any other of these kinds of accounts.
You can also check the local_user table to see if the email addresses are from the same provider (not a guaranteed way to match them, but it can be a clue) or if they're they same email address using plus-addressing (e.g. user+whatever@email.xyz, user+whatever2@emai.xyz, etc).
Why are they doing this?
Your guess is as good as mine, but US elections are in a few months, and I highly suspect some kind of interference campaign based on the volume of these that are being spun up and the content that's being manipulated. That, or someone, possibly even a ghost or an alien life form, really wants the impression of public opinion being on their side. Just because I don't know exactly why doesn't mean that something fishy isn't happening that other admins should be aware of.
Who are the known culprits?
These are ones fitting that pattern which have been identified. There are certainly more, but these have been positively identified. Some were omitted since they were more garden-variety "to win an argument" style manipulation.
These all seem to be part of a campaign. This list is by no means comprehensive, and if there are any false positives, I do apologize. I've tried to separate out the "garden variety" type from the ones suspected of being part of a campaign, but may have missed some.
[New: 9/18/2024]: https://thelemmy.club/u/fxgwxqdr
[New: 9/18/2024]: https://discuss.online/u/nyubznrw
[New: 9/18/2024]: https://thelemmy.club/u/ththygij
[New: 9/18/2024]: https://ttrpg.network/u/umwagkpn
[New: 9/18/2024]: https://lemdro.id/u/dybyzgnn
[New: 9/18/2024]: https://lemmy.cafe/u/evtmowdq
https://leminal.space/u/mpiaaqzq
https://lemy.lol/u/ihuklfle
https://lemy.lol/u/iltxlmlr
https://lemy.lol/u/szxabejt
https://lemy.lol/u/woyjtear
https://lemy.lol/u/jikuwwrq
https://lemy.lol/u/matkalla
https://lemmy.ca/u/vlnligvx
https://ttrpg.network/u/kmjsxpie
https://lemmings.world/u/ueosqnhy
https://lemmings.world/u/mx_myxlplyx
https://startrek.website/u/girlbpzj
https://startrek.website/u/iorxkrdu
https://lemy.lol/u/tjrwwiif
https://lemy.lol/u/gmbpjtmt
https://thelemmy.club/u/avlnfqko
https://lemmy.today/u/blmpaxlm
https://lemy.lol/u/xhivhquf
https://sh.itjust.works/u/ntiytakd
https://jlai.lu/u/rpxhldtm
https://sh.itjust.works/u/ynvzpcbn
https://lazysoci.al/u/sksgvypn
https://lemy.lol/u/xzowaikv
https://lemy.lol/u/yecwilqu
https://lemy.lol/u/hwbjkxly
https://lemy.lol/u/kafbmgsy
https://discuss.online/u/tcjqmgzd
https://thelemmy.club/u/vcnzovqk
https://lemy.lol/u/gqvnyvvz
https://lazysoci.al/u/shcimfi
https://lemy.lol/u/u0hc7r
https://startrek.website/u/uoisqaru
https://jlai.lu/u/dtxiuwdx
https://discuss.online/u/oxwquohe
https://thelemmy.club/u/iicnhcqx
https://lemmings.world/u/uzinumke
https://startrek.website/u/evuorban
https://thelemmy.club/u/dswaxohe
https://lemdro.id/u/efkntptt
https://lemy.lol/u/ozgaolvw
https://lemy.lol/u/knylgpdv
https://discuss.online/u/omnajmxc
https://lemmy.cafe/u/iankglbrdurvstw
https://lemmy.ca/u/awuochoj
https://leminal.space/u/tjrwwiif
https://lemy.lol/u/basjcgsz
https://lemy.lol/u/smkkzswd
https://lazysoci.al/u/qokpsqnw
https://lemy.lol/u/ncvahblj
https://ttrpg.network/u/hputoioz
https://lazysoci.al/u/lghikcpj
https://lemmy.ca/u/xnjaqbzs
https://lemy.lol/u/yonkz
Edit: If you see anyone from your instance on here, please please please verify before taking any action. I'm only able to cross-check these against the content my instance is aware of.
But this is SOO tedious. The annoying bit is it could just be one person who set it up over a weekend, has a script that they plug into when wanting to be a troll, and now all admins/mods have to do more work.
You're fighting the good fight! So annoying that folks are doing it on freaking lemmy.
I wonder if there's a way for admins to troll back. Like instead of banning the accounts, send them into a captcha loop with unsolvable or progressively harder captchas (or ones designed to poison captcha solving bots' training).
Not sure if shadowbanning can work here. Wasting each instance's limited pool of resources is not what we want to encourage.
Yeah not to mention it's not that hard to detect a shadowban if you're aware of the possibility. Lemmy doesn't even fuzz vote totals, so it would be trivial to verify whether or not votes are working.