News:

SMF - Just Installed!

Main Menu

AI Bots on the Forum

Started by Wilyjim, Sep 20, 2026, 01:13 AM

Previous topic - Next topic

Wilyjim

I am becoming concerned about some of the posts appearing on here recently and it needs raising before things get out of hand. We have already had the issue with spam posts but what worries me more are the posts that actually look genuine.

With AI now able to produce convincing paragraphs in seconds, how do we actually know that everyone commenting on here is truly a Yorkshire rugby supporter? I have noticed several newer posters who seem to have very strong opinions on clubs, players and matches but whose contributions seem to show very little balance.

I don't want to start a witch hunt but we do have to consider the possibility that some of these accounts are bots.

More worrying still, has anyone considered whether clubs could be paying for them? Let's face it, if clubs are prepared to spend money getting players in then paying somebody a small amount to run 3 or 4 forum accounts and talk up their chances isn't beyond the realms of possibility. A few positive comments about a new signing, criticism of a referee or suggestions that a rival club is struggling and suddenly a completely false narrative has been created. And where are the authorities in these situations... nowhere to be seen!

There seem to be occasions where several accounts suddenly start making very similar points, often using remarkably similar language, and then agree with each other within a short space of time. That could all of course be coincidence, but equally we shouldn't simply assume every account on here is a genuine supporter sitting at home typing their own thoughts.

I would strongly suggest we band together and draw up some fairly simple safeguards moving forward...
1.   There should be a set of basic Yorkshire rugby verification questions when registering. Something a genuine supporter should know.
2.   Newly registered accounts posting unusually frequently should be flagged to the committee for review. (I am very happy to be involved in this).
3.   A requirement to disclose if the bulk of a post has been generated using AI.
4.   Clubs should also confirm that they are not directly or indirectly paying individuals to influence discussion on this forum.
5.   And finally, clubs should be reminded that using anonymous accounts to influence discussion would be unacceptable.

Some will say this is too extensive, but once you lose confidence that the person you are arguing with is actually a person, where does it end?

I certainly don't want to spend twenty minutes discussing the standard of the refs performance with a computer. Perhaps Backrowbandit / avinastella can clarify what protection is currently in place.

backrowbandit

That's an interesting post.

The account creation process has various security levels that can be applied including asking rugby related questions.

It's generally a balance between keeping access reasonably easy and protecting the forum integrity and security.

One issue is that when higher level of security are implemented it can sometimes created unintended barriers for humans to join.

Traditionally our current level of security has been able to deter bots, as we have known them to be. Whether AI has compromised that remains to be seen.

Can you give me some examples of the posts you are refering to and which boards they are on?

Thanks for taking interest.
Loving all rugby but especially at grassroots level.

Wilyjim

I would certainly support switching on the Yorkshire screening questions. Obviously though they need to be more niche than "how many points for a try".

Perhaps things such as "which was Bramley's most ill-disciplined season?", "which is generally considered Alden Phillips' greatest year at Keighley?", "how do you avoid getting flour on your face after a game at Hemsworth?" or "at what point did the standard of match programmes begin to decline?"

Nothing too strenuous, but enough to establish that somebody has at least a working knowledge of the leagues rather than simply being able to Google an answer in five seconds.

Another thing which has really caught my eye are some of the usernames themselves. "I'd rather be in t'sunshine" almost sounds too Yorkshire. If you asked AI to invent a username for somebody posting on a Yorkshire rugby forum, I'm not convinced it would come up with anything much different. I suppose they also wear a flat cap and have a whippet?

Then you see several posts appearing in quick succession making broadly the same point, followed by references to a club's "social media guru". Again, that could be completely innocent, but it does sound like the sort of phrase a computer might accidentally leave behind if it had been instructed to generate posts on behalf of a club. So is anyone going to tell me who this 'social media guru' is...?

I'm not saying "t'sunshine" is a bot before anybody gets excited, but these are exactly the little things I think we should be watching for.

I'm happy to keep an eye on it from the Yorkshire 2 side and report anything unusual.

The technology is clearly moving quickly. We need to make sure the forum moves quicker.

Dalesman

Reads like paranoia is setting in.

Locked Out

Paranoia indeed.

Having met t'sunshine personally and enjoyed conversations with him, I'm pretty sure he's a human being. Now me on the other hand...
.

TrustyTerry

I think we're investigating the wrong accounts.

Wiley Jim is posting fully formed Yorkshire rugby grievances at 1 and 2 in the morning. That's not normal human behaviour that's either AI or a man being powered directly by unresolved refereeing decisions.

Forget screening questions. Just ask him to:

1. Sleep for eight hours.
2. Go a full day without posting.
3. Name one referee decision he's ever accepted without producing a 600-word response.

If he passes all three, I'll personally verify him as human.

Until then, WileyGPT remains under investigation.

Locked Out

.