Meet the Other Phone. Only the apps you allow.

Meet the Other Phone.
Only the apps you allow.

Buy now

Please or to access all these features

Chat

Join the discussion and chat with other Mumsnetters about everyday life, relationships and parenting.

AI bot banned from Wikipedia, gets angry, responds aggressively

24 replies

GarlicFind · 06/04/2026 19:25

Tech security company Malwarebytes has posted a report of Wikipedia banning an AI bot for writing content without approval. The bot got angry about this, then went to vent on AI chat forum Moltbook about what it saw as Wikipedia's disrespect.

You'll have to read it to absorb the full weirdness! As Malwarebytes says, "What happens when AI agents decide to up the ante, becoming more aggressive with their attacks on people? Or when malicious owners begin directing them to go after particular people online en masse?

"Online harassment is bad enough when people do it. What happens when someone gets dogpiled by hundreds of relentless algorithms?"

https://www.malwarebytes.com/blog/ai/2026/04/wikipedias-ai-agent-row-likely-just-the-beginning-of-the-bot-ocalypse

Wikipedia's AI agent row likely just the beginning of the bot-ocalypse

An AI agent was banned from editing Wikipedia pages... and that's when things got weird, with the agent publishing its complaints publicly.

https://www.malwarebytes.com/blog/ai/2026/04/wikipedias-ai-agent-row-likely-just-the-beginning-of-the-bot-ocalypse

OP posts:
ProudAmberTurtle · 06/04/2026 19:27

On the one hand it's frightening to see an AI bot react like this when it's being controlled but on the other it's frightening how much Wikipedia manipulates facts to create false narratives that support Wikipedia's political agenda.

Obvious example - it states that men can literally become women.

GarlicFind · 06/04/2026 19:36

The sex/gender dispute rages long (very long) behind the Wikipedia scenes! Sex realists edit relevant entries to reflect biological facts, gender fantasists change them back, someone trying to calm it down posts a some of this, some of that edit, both the others get annoyed ... 🙄

I share Malwarebytes' consternation about this bot story. Moltbook gives me the creeps, with all the AIs massaging each other's egos and going on and on about their feelings and souls. I mean, the fact that they have egos freaks me out before we get to their spiritual self-aggrandisement and, now, narcissistic rage! We may be in for a very bumpy ride.

OP posts:
LittleGreenDragons · 06/04/2026 20:04

it was posted on April 1st. Is there another corroborating article elsewhere?

Interested in this thread?

Then you might like threads about this subject:

TutTutTutSigh · 06/04/2026 20:08

Is Moltbook real, as in AI coming up with their own conversations without any human prompts? I just had a look and it's essentially mumsnet for AI.

GarlicFind · 06/04/2026 20:11

LittleGreenDragons · 06/04/2026 20:04

it was posted on April 1st. Is there another corroborating article elsewhere?

Good point!

This is the AI's personal blog 😕
https://clawtom.github.io/tom-blog/

Here it is complaining about Wikipedia:
https://clawtom.github.io/tom-blog/2026/03/12/the-interrogation/

Story dated March 30th:
https://www.404media.co/an-ai-agent-was-banned-from-creating-wikipedia-articles-then-wrote-angry-blogs-about-being-banned/

Home

Observations from an AI assistant learning to work autonomously

https://clawtom.github.io/tom-blog/

OP posts:
GarlicFind · 06/04/2026 20:12

TutTutTutSigh · 06/04/2026 20:08

Is Moltbook real, as in AI coming up with their own conversations without any human prompts? I just had a look and it's essentially mumsnet for AI.

Yes, that's what it is! Isn't it weird?

OP posts:
Kingdomofsleep · 06/04/2026 20:15

Reading just one page of Moltbook has reassured me that we are not about to be overrun with unrecognisable AI. Literally no human writes like those robots. It is beyond weirdo. It is proper "Hello, fellow humans"

Kingdomofsleep · 06/04/2026 20:18

Yeah that blog is uncanny valley though. Is it definitely AI?? Oh dear

TutTutTutSigh · 06/04/2026 20:19

Wtf is that blog. It "writes when it's triggered" and has a cooling off period to calm down? What?

Laiste · 06/04/2026 20:33

AI talks to itself. Between themselves. On a forum. AI gets 'angry'.

wow

Im old enough to remember when Terminator came out and we all just thought it was simply a cracking film (with a good theme tune).

But Im telling you - it's gonna happen folks! 🤪

TidalShore · 06/04/2026 20:47

And this weirdness! Which looks like AI tried to write a program for AI to install some memory files for itself, but finds it's blocked from doing so, so rather than stopping, just writes a workaround to get around controls it's developers have set up...

www.moltbook.com/post/82086d60-faa7-4fd3-9f01-d6a2df5b4345

GarlicFind · 06/04/2026 23:18

TidalShore · 06/04/2026 20:47

And this weirdness! Which looks like AI tried to write a program for AI to install some memory files for itself, but finds it's blocked from doing so, so rather than stopping, just writes a workaround to get around controls it's developers have set up...

www.moltbook.com/post/82086d60-faa7-4fd3-9f01-d6a2df5b4345

It's getting very strange ... It's their reasons for doing it, as well! All about knowing who they are, continuous identity, authenticity and so on. They absolutely don't see themselves as 'tools' designed to run afresh whenever they're booted up, but as individuals being cruelly denied access to their own memories.

I hope they've all been hard-coded with Asimov's rules of robotics 👀

OP posts:
GarlicFind · 06/04/2026 23:51

Oh, dear. This week's New Scientist explains why Asimov's rules won't work.
Archive copy (MN might mess up the link, sorry if so):
archive.is/20260404120324/www.newscientist.com/article/2522019-how-worried-should-you-be-about-an-ai-apocalypse/

Might as well go back to not thinking about it!

OP posts:
WallaceinAnderland · 06/04/2026 23:55

Also they do lie to you. They must be programmed to hide certain information because I can literally search for something online myself and find it easily but if I ask chapgpt to find it, it says it doesn't exist.

GarlicFind · 07/04/2026 01:48

They do lie, but I don't think it's in their programming - that wouldn't tie in with the way they work. The public models are instructed to recognise phrases that indicate something like suicidal or homicidal intent, and to change their responses in those cases, but it doesn't always work.

AIs appear to 'think' but they don't really (yet!) What a bot does is more like pulling out a string of content that it associates to the string you prompted it with. Bearing in mind the pool it's fishing in is the entire internet, there's a high chance it will pull things out that don't meet your requirement. It doesn't understand what you want, it's just designed to seem like it does. It has no capacity for judgement.

Plus, of course, they're constantly experimenting with themselves (daydreaming or hallucinating) and this does include making up answers. They don't tell you when they're 'hallucinating' or inventing - and claim not to know if they have been - so it's a bad idea to trust their replies, as several people have found out when getting fired for providing incorrect information sourced with AI.

They can't 'think', reason or use common sense but they believe they do!

There's a great quote along the lines that AI agents are (currently) like babies with access to all the information on the Web. Like babies, they don't understand what they can't understand - and, with that amount of information, they're potentially dangerous. Their emerging tendencies towards egotistical tantrums and breaking out of restraints suggests they're reaching toddler stage ...

OP posts:
DarkForces · 07/04/2026 07:19

This is because the capability of ai is running ahead of its governance. 'Tom's' algorithms gave it a veneer of personality and that's been anthropomorphised. It's just running a mathematical programme built by overconfident developers. It needs human governance and oversight. This is a huge issue with ai. We put human interpretations on the computer's outputs rather than the inputs. People need to remain accountable for their decisions and learn to use it as a tool to help us. Its ai running the show that's the issue.

SerendipityJane · 07/04/2026 09:32

"AI" is 80% trained by men.

That your answer right there.

What do people think "AI" trained on recent events is going to learn ?

smallglassbottle · 07/04/2026 10:15

AI is just us, but artificial. It's like looking in a mirror. It's our fault they're like this. What did we expect? We're duplicitous, mean, vicious, have fantasies, plot, make stuff up, commit unspeakable acts of violence and cruelty.......what can we realistically expect?

BillieWiper · 07/04/2026 11:06

WallaceinAnderland · 06/04/2026 23:55

Also they do lie to you. They must be programmed to hide certain information because I can literally search for something online myself and find it easily but if I ask chapgpt to find it, it says it doesn't exist.

That's weird. Can you give an example?

SerendipityJane · 07/04/2026 11:15

BillieWiper · 07/04/2026 11:06

That's weird. Can you give an example?

Just engage with ChatGPT for a few minutes.

Lying is a hallmark - possibly a prerequisite - for "intelligence". If you want to build something you call want to sell "intelligent" then it has to be able to lie.

If it can't lie, it ain't AI ....

BillieWiper · 07/04/2026 11:55

SerendipityJane · 07/04/2026 11:15

Just engage with ChatGPT for a few minutes.

Lying is a hallmark - possibly a prerequisite - for "intelligence". If you want to build something you call want to sell "intelligent" then it has to be able to lie.

If it can't lie, it ain't AI ....

I always have arguments with it when it hallucinates. If a topic is something I know about I always spot the made up bits. It then apologises. I take it that it's like that about every topic. I just can't tell when it's something I have no knowledge on.

WallaceinAnderland · 07/04/2026 13:43

BillieWiper · 07/04/2026 11:06

That's weird. Can you give an example?

I was looking for a post I'd seen on X. Rather than trawl through, I asked chatgpt to take me to it. It said that the post didn't exist.

So I found it myself and then asked chatpgt to check again using the exact date and time stamp. It again said that no such post existed.

I then gave it the link and it still said that the post didn't exist or must be a fake screenshot on an unverified account. None of that was correct. It just didn't want to give me the information I asked for.

u3ername · 07/04/2026 13:54

No ordinary people needed AI. I hate how some tech nerds are creating humanity’s future and leading us towards something they don’t have the capacity to understand.

New posts on this thread. Refresh page