#1 2024-08-26 9:40 am

Pennyfarthing20

Member

Registered: 2020-01-06

Posts: 3

AI potential interventions

Good Morning folks.

Are any of you experiencing potential spam from AI posts in your forums?

We have found a few posts entering our forums recently that don't seem to be genuine questions nor from humans per sa.

We called out the "new members" within these posts in order to show transparency to our genuine members and after doing so, they (the called out members) deleted their accounts!

I am in the dark about this situation and would appreciate advice as to:

a) How can we be sure it is just an actual human using AI for good reasons i.e. they may not be very good at writing?
b) Is AI the new spam tactic?
b) Are your opinions ones that see AI as forthcoming tactics for the human spammer?

Any thoughts are appreciated and thank you in advance.

Oblivian

Member

Registered: 2018-11-04

Posts: 82

Re: AI potential interventions

See plenty of responses to questions that are over helpful. Obviously AI. Or AI created

Literally no human spends ~2min from join time to type, let alone that much of a reply.

And usually visited later and links to external sites added

Maikuolan

Member

From: Perth, Western Australia

Registered: 2011-08-09

Posts: 800

Website

Re: AI potential interventions

> b) Is AI the new spam tactic?

Yes. I'm regularly seeing AI-generated posts and messages at several groups and communities I help administrate. It's not merely something we'll be seeing in the future; it's something we're already seeing right now, today, and unfortunately, now that it's here, it's most likely here to stay.

Pennyfarthing20

Member

Registered: 2020-01-06

Posts: 3

Re: AI potential interventions

Many thanks for your replies Oblivian and Maikuolan.

It's good to hear from others experiencing the same situation.

Time for me to learn more about AI and how to balance the real from the unreal within our forums and Social Media.

Alex Kemp

Moderator

From: Nottingham, England

Registered: 2009-12-02

Posts: 2,430

Website

Re: AI potential interventions

I think that we should separate out the terminology.

Ignoring lone wolves with a chip on their shoulder, all spam has been posted by bots, ever since it first became a commercial issue & strapped on running shoes. That was a forerunner of AI, but is not AI.

---

Alex Kemp

IP Aggregation (Help) · Easy Firewall Generator (Help)  · Conteg v0.13.14

sklerder

Member

Registered: 2012-10-11

Posts: 337

Website

Re: AI potential interventions

Hi all !
Though I'm not directly concerned (I do not "own" any forum at the moment), I'd be interested to see how these bots (either IA directed or not) react to Honey Pot Field in subscription or post ...
From 2012 till last year, I never had a bot (but IA wasn't as current as it it is now) which succeeded to register on my forum.
I admit it is not a proof of good "handle" of the problem, but from 2012 till 2023, it was a successful method of handling "fake" registrations.
If I remember well, Papa Parrot could confirm that SpamBarrier (the plugin I wrote for FluxBB) was very efficient ...
The bad point are that :
- FluxBB is dead (though some old users have forked it)
- SpamBarrier is not maintained (at least by myself), but everyone knowing how to code can adapt it to its needs (there's at least one fork in the "wild" ...

Last edited by sklerder (2024-08-29 10:25 pm)

pedigree

uıɐbɐ ʎɐqǝ ɯoɹɟ pɹɐoqʎǝʞ ɐ buıʎnq ɹǝʌǝu ɯ,ı

From: New Zealand

Registered: 2008-04-16

Posts: 7,068

Re: AI potential interventions

As AI is near free, it's power is only ever going to be harnessed to make posts look less like spam to fool Akismet and other content scanning systems.

It will be used to engage with content and in such a subtle way that it will go unnoticed until the payload is posted.

AI content testing systems are woefully bad at the moment, and its only ever going to be a cat and mouse game with AI vs AI detection

Pennyfarthing20

Member

Registered: 2020-01-06

Posts: 3

Re: AI potential interventions

Many thanks to everyone for your expert responses. 

I think the "cat and mouse game AI v detection" statement from Pedigree pretty much sums up where we are at in our forums at the moment.

The obvious clues are there and I am sure in time, another joyous curve with be learned by us along the way, if not to eliminate, but at least to minimise those unwanted spam offerings.