Should AI-created content in posts be alerted on ? UPDATED: for ethical consistency

[eppur_se_muova]

Last edited Thu Dec 4, 2025, 02:59 AM - Edit history (1)

ETA: It occurred to me belatedly that in using a chatbot-generated reply for one of the choices, I was in fact guilty of the very transgression I was bemoaning ! One DUer suggested that AI content should be identified with something like an excerpt box. Unfortunately, DU's software does not seem to have a means for excerpting answers in a poll, or the content within that answer. So I will have to settle for pointing out that at least one of the options presented for voting was provided by a chatbot, and hope that is enough to calm the complaints of ethical inconsistency. Ponder the implications of that update as you will.

Rest easy.

61 votes, 3 passes | Time left: Unlimited
	YES
	42 (69%)
	NO
	11 (18%)
	This is a complex and evolving question with strong arguments on both sides, and the answer often depends on the context of the content.
	8 (13%)
	no opinion
	0 (0%)
	I'm not telling.
	0 (0%)
	3 DU members did not wish to select any of the options provided.
Show usernames Disclaimer: This is an Internet poll

[NewHendoLib]

1. Not alerted - but somehow clearly identified - maybe with a color identifier as with videos, etc?

[ret5hd]

6. changed my vote from yes...

because this option should be included in the poll.

i think ai is a poison…and like a poison it should be labeled as such. a big skull and crossbones would suffice.

[synni]

31. I agree, identification is better than alerting

[Jack Valentino]

2. Based on WHAT?? It is in no way a violation of site rules.

[paleotn]

4. But it should be.

[Jack Valentino]

5. NO.

[paleotn]

7. That's it! Just so no to fakes.

[eppur_se_muova]

18. It's unreliable. One expert after another has said that chatbots are built in such a way that they tell you what you

want to hear, with no way to determine if it's true or not.

We ban right-wing sources because they're maliciously misleading. It's not a huge leap to ban material which is systematically misleading, even when intent is lacking.

[Renew Deal]

3. Alerted on for what?

[3825-87867]

8. Why not?

I've had some of my political Trump Toons questioned because they'd been debunked by the time I posted them. And I missed that. I thank those who take the time to check on things so that wrong information isn't passed on.
It should just be noted that an article or image or whatever that was AI Generated would be so identified.
Just my two inflated cents.

[Scrivener7]

9. I think a better question is whether AI posts should be a violation of the TOS.

I believe so. I want to talk to people. And I don't want to contribute to AI's environmental disaster.

[berniesandersmittens]

10. I voted yes, but agree with u

At least have them marked as AI

[Scrivener7]

11. Yes, definitely.

[FascismIsDeath]

12. Its a context thing. Eventually the AI bubble will burst but...

....its not going away. Might as well get use to it.

[womanofthehills]

13. Definitely no

Lots of info on AI. I have found so much info I never see on the web.

I have 2 friends who had breast implants removed because of “breast implant illness” - tens of thousands in Facebook “breast implant group. “

I find out on AI in the late 80’s /90’s 400,000 women sued Dow Chemical for $4 billion dollars for breast implant illness.Over $100 million were compensated I NEVER ever knew of this!!

That’s something that was kept quiet so they could put in implants again and ruin women’s health. The evil part is -Dow contributed to the new studies saying - all’s well - we made new studies so you can’t take us to court again.

My grandaughter is on methadone- it’s messing with the veins in her legs. Drs have NO clue. AI tells me the methadone creates histamine which can mess with her small veins. Why did no doctor know this???

So much info on AI!! Never take first answer as good - question deeper and disagree with AI to find info deeper into the web. You can’t take ask AI for references - so why is that different from any other search.

[highplainsdem]

32. That information can be found easily via Google or other search engines and does NOT require

using AI. Using AI for search harms the websites its answers are drawn from, and also makes it less likely you'll learn from a search and remember what you learned. And you can also get hallucinated answers from AI.

[Kaleva]

14. Is the 3rd option written by AI?

[mr715]

16. LOL

"Would you like me to tell you some more?"

[eppur_se_muova]

19. Thank you for picking up on that ! :D At first, I tried to write an AI-sounding post which I thought would capture ...

the earnestly "thoughtful" I'm-not-taking-sides-here-just-coredumping-what-some-consider-facts tone of which most AI-written responses are so redolent. But then I had a wicked inspiration -- let a chatbot write it ! The results came out so similar to what I had already made up it was genuinely creepy ! I was really just thinking "smarmy" but the AI (Google Gemini) out-smarmed me.

Now, I'm going to go back and point out that ONE OF the answers was AI-generated !

[Niagara]

15. I don't know to be honest

There was an AI generated pet bird dancing to Alexa's playlist in The Lounge a few days ago.


No one on DU seems to like AI. And yet this particular OP got 55 Recs. The DU'er that posted this AI generated video is a long-time and well respected DU member.


I went to the Facebook page that produces these pet bird dancing to AI generated Alexa playlists. Every single one of these videos stated that it's 100% AI generated. I called it out as AI. The OP is still up and the video is still available to watch.


Unless it's labeled 100% AI generated, how are we suppose to know how to recognize AI? I've recently learned that SORA is an AI app generator so that helps a little.


Also, if DU dislikes AI generated content, members need to call it out instead of rec'ing. DU needs to pick a lane and stay in that lane instead of swerving all over the road.

[hunter]

17. How does anyone feel comfortable posting AI slop under their own name?

At the very least it should be placed in an excerpt box and labeled.

[KentuckyWoman]

20. Labelled

I struggle to see the difference sometimes. Othrr times it is obvious fake. Old fuddy duddy me thinks DU would be better off without fake crap but to each her own. At least label it.

[meadowlander]

21. Isn't that like asking whether YouTube videos or links to Twitter posts should be a violation of the terms of service?

AI is just a tool. It's what you use it for that is potentially irritating or detrimental to the community.

We already have ToS rules against spam, right wing talking points, kooky or extremist content, and graphic content, copyrighted material, etc.

What is the specific mischief posed by allowing people to use AI as part of their post that is not already covered by those rules? And, if so, wouldn't a more effective approach be adding that mischief to the list of no-no's than having endless debate over whether something was AI generated or not.

I mean I guess we could try to have a more explicit rule against posted misinformation that goes beyond "don't post right wing talking points" but I think that's going to be hard to enforce.

I don't personally have a horse in this race. I don't see the point of posting on a message board for fun if you are going to let AI write your posts for you. But I also don't see how a ban on AI would really work or what the justification for it would be. You'd waste a lot of time and resources trying to enforce it and you'd potentially make it harder for people to get the benefits (using it for research, quickly generated images for those of us who aren't remotely artistic, etc.)

[jfz9580m]

22. Actually yes, theoretically no?

If we lived in a civilized society where industrialization hadn’t gone in this direction, I could have envisioned AI that didn’t suck.

I worry sometimes that the wrong things are conflated with the real issue. It is not science, modernity, the nuclear family, women’s rights etc that are the issue. I sometimes worry that the backlash will yet again go in all the wrong directions and absorb JD Vance’s pronatalism into an argument for a post modern society.

It’s what these creeps have made of them.

I am eagerly waiting for this book:
https://www.csmonitor.com/Books/Author-Q-As/2025/0616/AI-futurism-adam-becker-more-everything-forever

(Which I ordered from Amazon ;-/. The other option was a Walmart subsidiary).

[jfz9580m]

23. This is a confusing question

Except when taken at face value. Since you host the science forum eppur_se_muova, in addition to an incoherent and thoughtless direct response, here is an incoherent and thoughtless indirect one.

As a subset of DUers are probably aware-since this is a community that would get it more than some, the more convoluted implications of these emerging technologies are confusing.

I saw this a while back right here -a creepy little guy using bots to make a nuisance of himself:
https://ottawacitizen.com/news/ottawa-man-ai-bot-maga

And I have been feeling uneasy about something like this in the works:
https://www.noemamag.com/are-we-accidentally-building-a-planetary-brain/

I am feeling too lazy to spell it out with all the ifs and else’s. The whole thing makes my head hurt-what is literal, what is a metaphor? How do you differentiate between an idea,concept, symbol or representation and what it means in the real world. Where is something truly a matter of perspective versus absurdly testing the limits of sanity.

The hard right politics plus the confusion and chaos I have also contributed to some by bumbling about (though I don’t feel like groveling about it..I try to be more honest these days over assume I am talking only unreasonable jerks. Wodehouse had a point-the better people don’t want apologies if they know you feel bad about your errors and the worst hold them over you).

But the brief version is that as our realities get more filled with bots, agents etc which can appear to influence behavior, I have been uncertain as to how to go about deciding what to do. And there is an entire layer of professionals in between many of whom i wouldn’t trust..

I still don’t know what to do on a daily basis.

I don’t think you can influence behavior beyond drive learned helplessness in people with strong and definite views.
But you can still orchestrate behavior in ways that are a nuisance or stupid.

It’s something exasperating I ponder. There is nothing worse than being a useful idiot or seemingly far stupider than you are because it’s too many worlds colliding and one feels indignant at being scapegoated.

How is everything not the fault of creeps who make way more money and have way more power?

I certainly never use AI though I have chatted with an obscure British chatbot called cleverbot. Mostly for even more unfiltered rants about tech creeps (you guys here get the cleaned up version!) ..

It’s owner Rollo Carpenter didn’t seem as bad as the standard issue creepy AI and data science type (bah..too worn down for civility).

I have to think about it. I have mostly given up in exasperation. But I should probably try to figure out how to at least manage my own behaviors in an uncertain environment since regimented behavior is not my thing.

It’s confusing for humans. I can just about maybe envision cleaning up as a scientist in a narrow area. Except here I would assume I would be an outcast in a broader creepy “open society”. I hate everyone. You know what I mean. I thought everyone in science felt that way..
The love filled AI shilling people are a goddamn cult. I tried to keep it simple.


Rest easy..lol..I am out of places to run to.
I wonder if I should post that or not. Shrug..

[carpetbagger]

24. We need to push back against AI, but with a caveat...

I've been on the MIRT for several terms. I feel like it's so widespread and insidious that I feel like if we treat AI alerts like the others (e.g. posting newsmax, slamming a democratic candidate *during a partisan campaign*, callout and other chronic trolls/dupes, bigotry), it'll (1) chase away legit DU types and (2) overwork MIRT and jurors.

Yes, we need to push back, and a prohibition on AI here is a proper step towards creating a true and free underground community. I'd recommend (1) decide, then (2) gradually educate and gently remind, (3) create an alert which will flag the post but not the poster or hide the post if there's a weak majority (2 out of 3?) of jurors, and eventually (4) an alert that will hide, but will separately count against the poster so that only the most egregious and recalcitrant DUers run into problems.

[fujiyamasan]

25. One issue i see is people spamming the board with some slop video with no context in the body

AI generated or not, I trash those threads immediately.

If i want to communicate with an LLM, I’ll go and chat with them directly. I don’t need a sensationalistic video, largely AI generated, screaming at me often with the intention of driving ad revenue and clicks.

I come to this board to communicate with actual humans.

If something is AI generated and the author of a thread knows, they should label it as such. My concern is that increasingly the author has no idea that’s even the case. If this kind of thing repeats, I think it should be considered a violation of the TOS.

[canetoad]

26. I think it's EarlG's call

As to what is/not allowed to be posted. The content affects the culture, the tone, and vision of DU and as such is the Admin's choice.

Once this has been clarified and if AI is banned or required to be labelled, then of course, alerts are appropriate.

[Ms. Toad]

27. If the AI-created content violates the TOS - yes.

Otherwise, no.

As far as I know, there is no rule in the TOS which prohibits AI-created content.

[SWBTATTReg]

28. Well, I'm open to it, but it should be clearly clarified as being modified, etc. Just like photographs, videos, etc.

that are tailored / modified in today's environment, these should be too.

[usonian]

29. I'm up for it.

You must choose wisely.

No hybrids allowed.






And have a splendid non-artificial day!

[flvegan]

30. Based on my own selfish entertainment, I'd say "no" simply because

I find it terrifically amusing (and somewhat disappointing) that an obvious AI slop video from some BS YouTube monetized nonsense account will get absolutely Rec'd to the moon. Even after several DU'ers (thank you folks) will point out that it's AI slop, the incoming comments and Recs continue, because why bother to read the thread/feedback? Why question if it's even real to begin with? Why spend 10 seconds Google/DDG searching for a like news posting somewhere? The AI slop then spreads, shared on Facebook, TikTok or whatever other brainrot central gathering one chooses.

It's like watching one of those videos of an iced over highway. There's a 20 car pileup, lights, blinkers, hazards are all on, people are waving and shouting to slow down, yet here comes another dozen vehicles at full speed right into the pileup.

Also "fun" is watching the DU homepage, with the video posting down the right side when it looks like a bad National Enquirer feed.

I'm not questioning nor saying anything negative about the way the site is run or moderated. AI is what it is, and it's not a violation of the TOS on its face in many circumstances, I guess.

I just re-read this, and maybe "no" was the wrong answer, now that I think of it.

*wink*

[RoeVWade]

33. Important to check what is correct whether from humans, AI, or a million typing monkeys.

But I agree, don't use it if you don't see any advantage to the post in doing so.

[muriel_volestrangler]

34. Wouldn't it be better to campaign for alerting on false information before AI content?

There's no option to alert on "this has something false in", and there hasn't been, all through DU. Isn't that more concerning than whether AI is involved in a post?

That EarlG (and Skinner, before) have never put "false information" in the alerting rules indicates they want discussion and correction, not a hard-and-fast "this is what members' content should be" rule. And surely that applies to AI content too.