YSK: Someone is creating sockpuppet account after sockpuppet account self-promoting an AI slop plagiarized news site.
They're trying really hard to make their AI-plagiarized slop "news" site happen. You'll see posts linking to primestories24 [dot] com but they're AI slop regurgitating likely legit news articles. Every "reporter" has some generic profile that is most likely the system prompt used for everything they "write".
If you see a post to that domain, report it as spam. The accounts all look similar to this:
- @[email protected]
- @[email protected]
- @[email protected]
- @[email protected]
- @[email protected]
- @[email protected]
- @[email protected] (First ban-evasion alt)
- @[email protected] (Original account)
This list isn't comprehensive.
Update: They are now using sockpuppet alts to vote/boost their posts. Just found @[email protected] that way and, surprise surprise, it was registered the same time as the latest batch.
Note: Technically the above is violating rule 7 if you consider spambots to be members. I don't, but if this is a problem, let me know, and I'll remove that part.
The domain was registered in May of 2026 and is roughly when these accounts started popping up.
Domain Name: PRIMESTORIES24.COM
Registry Domain ID: 3105727536_DOMAIN_COM-VRSN
Registrar WHOIS Server: whois.web4africa.net
Registrar URL: http://hostafrica.com/
Updated Date: 2026-06-19T10:01:29Z
Creation Date: 2026-05-31T12:16:00Z
Registry Expiry Date: 2027-05-31T12:16:00Z
Registrar: Host Africa (Pty.) Ltd.
Registrar IANA ID: 664
Registrar Abuse Contact Email: [email protected]
Registrar Abuse Contact Phone: +1.6465850088
Domain Status: clientTransferProhibited https://icann.org/epp#clientTransferProhibited
Name Server: DAN1.HOST-WW.NET
Name Server: DAN2.HOST-WW.NET
DNSSEC: unsigned
Admins, I'm aware Lemmy is severely lacking in moderation capability, but one thing it does have is domain blocking. That prevents posting to those domains and it now seems to prevent links to those domains from federating in.
42 replies
Wow. PrimeStories24. What a trustable name! Registered on May 31st. Not even a month ago…
Oh man, it would be so cool if someone connected that to an SDR setup!
https://www.thedrive.com/news/every-gta-in-game-radio-station-is-playable-irl-in-this-guys-car
If you haven't taken this directly to instance admins, you probably should do that too. Not necessarily World's, if they're not posting to it, but definitely the ones on which they have made accounts.
Yup, I’ve noticed this over the last few days too. Reported and blocked each account I’ve found.
OP, you can pull more WHOIS data from the whois.web4africa.net WHOIS server for that domain. You only did a single-level WHOIS lookup to the North American WHOIS server; whatever WHOIS client you used doesn't default to recurse.
This does fit with the other material I commented here referencing an "Olawale Israel Ashorobi".
REDACTED FOR PRIVACY
I bet that's just the name they entered for the organization. Does it even have to be tied to anything real or do DNS registrars just accept any string for that field?
Having registered plenty of domains I can tell you that it needs a valid address but that address can be some random kebab shop the other side of the country, they don't care.
The address is on street view. Dude lives in a mansion behind a barbed wire fence lol.
Wonder who is paying for that mansion.
edit: whole streetfull of these homes.
Of course Israel is behind it again
That domain is now blocked on lemmy.ca and piefed.ca
Just in general I've noticed an uptick in posts that link to a news article of sorts without having anything else included in the post. Just a clickbait title and a link, likely to draw traffic to their website. I'm getting really sick of finding an interesting post just to be met with a "Subscribe to view article" message instead.
probably the same kind of spammers that also does this on reddit. i was a forum where they have discussing driving traffic through the site by using methods to just spam on multiple threads/instances.
Wouldn't be the first to succeed with trash. There used to be a German news aggregator site, shortnews.de, where some asshat was also permanently submitting links to their own site (and somehow the mods let them despite linking to your own sites was not allowed per the site rules), and while shortnews is long gone, that idiot and their trash site actually made a carreer out of it and they are still around :(
Which site?
I avoided naming them to not give them exposure. But if you must know, it's "promiflash"
Can Lemmy ban outgoing link URLs? Given the spread of slop websites, that seems like useful functionality.. Could even combine it with an AI slop URL blacklist...
Edit: Apparently it can do URLs and not just domains. See @[email protected] 's reply below.
URLS, no, but it can disallow specific domains. So it could work against specific spam domains but not against a bunch of accounts spamming, say, a Github repo, unless you want to block links to all of Github which isn't ideal.There's a list of disallowed domains that can be configured in the admin area. Until relatively recently, it would only prevent users of that instance from posting to them, but somewhat recently it now prevents inbound federation of anything linking to those. If something links to a blocked URL (post, comment, user with that in their bio, etc), Lemmy will reject it.I don't know why more admins don't look at the spam that gets modded and add the domain to their URL block list other than it being tedious work lol.
Seems like scripting that wouldn't be a super heavy lift.
It does work on URLs. The https part gets stipped out because of how the regex is built internally, but you can absolutely block a Github repo. I have a Discord invite link in the feddit.uk one.
It always worked on federated content? It was added in response to a Mastodon spam wave (source: I added it).
Oh, nice. For some reason I assumed it was only domains.
I know it has for a good while but I forget when it was applied to inbound federation. I think it was that Mastodon/Misskey spam wave I was thinking of but don't recall which version was running at the time and how I wished it would stop content to those domains from federating in. Nice contribution, BTW.
Thanks! I have also floated the idea of a regex blocklist, but Nutomic shot that down because 'it can be done with a plugin'. Can't wait until 1.0 so Lemmy can have feature parity with a 2006 forum.
I think that a heavyweight automod feature that works something like SpamAssassin, where one builds a training corpus of known spam and known ham, and then one writes a bunch of rules (e.g. "domain is less than X days old", "account has less than N votes", etc) and then runs it on the training corpus to auto-generate score weights for the rules, and then runs the rules on each post to provide an estimated "spam probability score" on given posts would probably scale better.
Doesn't need to be built into Lemmy or Mbin or PieFed, and running as an automod bot, it'd be able to run on all of those platforms, share effort among them.
I don't disagree, and the plugin system is cool as it'll enable people to build things like that. I think SJW even has something like that already. But I can't begin to tell you how much work a simple text filter would have saved me as an admin.
There were various Reddit automod bots that went after these, but then you had people trying to evade them by using one of the zillions of link shortener sites out there, so then you also have to ban link shorteners or try to resolve all the links (which may be a good idea as well, but saying that it's a more-complicated problem than initially meets the eye).
Also, I was banned in a few communities by an automod bot on the Threadiverse, and given that it happened when I posted some comment linking to a number of sites, I assume that one of the URLs must have tripped some automod list, so I'm a little irritable about bots running permabans on blacklisted sites, as users can inadvertently link to them.
I'm of a mind that link shorteners should be banned everywhere. You should never click a link unless you know exactly where it's going. Plus, some of them have interstitial ads and/or get deleted after a period of time, so even in the best case scenarios, they're just spraying out ads and link rot when you get down to it.
Not to mention, they're almost always used on platforms where you can just click on the link regardless of how long and/or unwieldy it is.
The only use case for bit [dot] ly and their ilk are if you need to publish a long link in print medium (such as a newspaper public notice or something) where a full link to a specific page would be unreasonable to expect people to type into a browser by hand.
The primestories24.com HTML source and page text contains multiple references to a "ZWadTech" and to a [email protected] email address.
Google's AI Overview on that domain --- no idea what data they pull from for it, so I'd take it with a grain of salt:
This "Selar Storefront" page also references the same name:
EDIT: As I said above, I'd be inclined to take the above Google AI summary with a grain of salt. While the Mastodon StoryLands account is directly associated with primestories24, and posts Christian stuff, and there is an Olawale Israel Ashorobi who has a Christian book out, it looks like there are a number of "Olawale Israel"s out there, and I have no idea what sources of information Google AI combined to generate the above; it may be merging multiple people.
EDIT2: Also, because I kind of feel that it'd only be responsible to mention this if digging up information like this: remember that identities may be real, but that people can also potentially steal identities. Those of you who recall the Nicole saga here...someone apparently was attempting to impersonate a real person, pretending to be a "Nicole", and then spam many users with direct messages. I understand that some of the admins on the Threadiverse looked into it, and believed that this was an attempt to get users here angry and harassing that person, who was not actually the person sending the messages. While I rather doubt that that is the case here, I'd suggest that going and trying to make this Nigerian guy miserable via attacking him or his stuff may be counterproductive. I'm posting the stuff because it could help filter spam stuff out, not as a call to arms.
More accounts:
https://mastodon.social/@StoryLands
https://www.youtube.com/@PrimeStories24news
Other social media accounts that also link to the site:
https://www.tumblr.com/primestories24
https://www.tiktok.com/@primestories24
You might also consider:
Yeah, I mentioned the list wasn't comprehensive and was just an example of the username format they're using.
Replying to your other comments here, but good deeper dive. The abridged whois was sufficient to confirm pretty much all I needed to confirm (that it was one of those pop-up AI plagiarism lookalike news sites) and that there was an active campaign to promote it here. Since I don't use any other social media, I was just reporting the ones flooding the Threadiverse but definitely seems they're flooding all the social media zones. I can't do anything about those, but figured I'd throw out a warning here.
Even more accounts:
https://t.me/primestories24
https://x.com/Olawale_gram321
Note that the X account has a profile picture that appears to be derived from the profile picture on the Facebook account, and a matching name.
Thanks for the heads up. Good looking out
Thank you for your diligence Admiral
Searching Google for "primestories24" turns up:
https://www.facebook.com/PrimeStories.24/
This page also links to an URL on that website about Janet Jackson, as well as some others, so I'm assuming that this Facebook username is associated with the site.
https://primestories24.com/according-to-singer-amaarae-janet-jackson-is-the-true-king-of-pop-not-michael-jackson
Note that the profile picture there is used to create the X profile picture:
Assuming that that image is not itself AI-generated, might actually be the spammer.
DavrielleLouna moment
DL/Dwazou operates (operated?) like a spambot cranked up to 11 but at least they posted legit stuff. These are just slinging the same AI-plagiarized slop site from a bunch of different accounts.
I see your account is a similar age to my own, so you may remember the infotinkerviral spam slop that was going around a few years ago. Not sure if I spelled that right, but similar M.O. to that.
i think i do yeah, you're right that the way they spammed is different