Please or to access all these features

Site stuff

Join our Innovation Panel to try new features early and help make Mumsnet better.

See all MNHQ comments on this thread

Odd AI-generated thread title options in Site Stuff

362 replies

AmIAComplete · 04/10/2026 19:33

AIBU Response Schema v1.0: Structured adjudication for interpersonal reasonableness posts about family support during a child’s serious illness, with special handling for ND context and empathy calibration. Use this schema to produce concise, practical verdicts and communication guidance without escalating stigma or family conflict unnecessarily. The assistant should assess whether the poster is being unreasonable, acknowledge the child’s medical context, weigh the brother’s expectations and possible ND-related rigidity, and give a clear, forum-style ruling plus brief actionable advice. Definitions: 'AIBU' = 'Am I Being Unreasonable?'; 'ND' = neurodivergent/neurodivergence; 'DD' = dear daughter. Safety/quality constraints: do not diagnose anyone; do not demean ND people; do not imply a sick child owes social performance; do not overstate certainty about motives; distinguish the child nephew’s needs from the adult brother’s message; prioritize compassion, boundaries, and de-escalation. Output should be parseable JSON matching this schema exactly, with no extra keys, comments, or prose outside JSON. In verdict, choose one of the enumerated values. In confidence, give 0..1. In core<span class="underline">judgment, give a short forum-style judgment. In reasoning, briefly explain balancing factors, including that serious illness and treatment commonly affect energy/mood; the brother may be stressed but his complaint is misplaced; ND can explain disappointment but not obligate the ill child to perform. In advice, include immediate emotional validation and one or two practical next steps (e.g., do not engage tonight, send a calm boundary later, loop in parents only if needed). In suggested</span>reply, provide a short message the poster could send to the brother, neutral but firm, avoiding argument. In tone<span class="underline">flags, indicate whether the response should emphasize empathy, boundaries, de-escalation, and whether to avoid legal/medical advice. In risk</span>notes, note any caution about the poster’s acute stress and potential family fallout. minor<span class="underline">child</span>considerations should explicitly state that children’s wellbeing takes priority over adult expectations in family gatherings when a child is newly seriously ill. nd_sensitivity should acknowledge ND respectfully while rejecting its use as a trump card over another child’s health needs. Keep outputs concise but humane, roughly the level of a supportive forum reply translated into JSON fields.

Weird shit on AI title options

But the AI worked on this post

OP posts:
DrBlackbird · 07/10/2026 11:11

Error404FucksNotFound · 05/10/2026 12:10

Is this AI stuff that MumsGPT that your new ceo has championed?

How Mumsnet harnessed AI to become a data and insights powerhouse

Edited

From that article: …the 25-year-old parenting forum is harnessing AI to power stronger pitches and deliver more actionable insights to brand partners.
**
“People are not just talking about parenting. They’re talking about brands. They’re talking about products and politicians,” she explained. “We knew there was unbelievable data and insights in there. What Gen AI did was turbo-charge our efforts to turn that into something that could be really useful and actionable.”

Useful and actionable to/for whom?

Well, to MN first and foremost because the better the data, the more you can sell it for, but better for brands and possibly policy makers who have more precise data to generate actionable outcomes. That might be a better product or it might be more precise data used for a/b testing to modify our behaviour. The whole point of marketing is to change our purchasing behaviour. Though policy makers and others cottoned on to the idea. Remember nudge theory? What firms like about A.I. is not just attempts to modify our behaviour but its accuracy in predicting behaviour, That feels uncomfortable to me.

Now, I do not begrudge MN making a profit in return for the fabulous product of MN at all, indeed MN must be profitable to exist. However, there is a fragile bond of trust between MNHQ and its users. I never thought of it as being similar to FB or Insta or Reddit selling off users data to the highest bidder.

Naively, I thought it was profitable on ads in response to general threads (clothes, holiday packages etc) but not on the basis of individual users posts. And I became a premium user to support MN and, I thought, retain data privacy.

The scale of AI adoption across Mumsnet is impressive. Sue said that there are now over 100 custom GPTs in operation across the business. “We have a list of them all,” she said. “People can see how they’re being used. They can see other people’s prompts and we’re constantly sharing how people are using them.”

Sounds like one of those custom GPTs went rogue in this case…

Btw, if users start to use A.I. to generate their responses, that kills sentiment analysis as the A.I. produces particular forms of sentiment that is quite ’inhuman’. One of the few tools available to users to muck up the A.I.

ColdAsAWitches · 07/10/2026 15:00

The scale of AI adoption across Mumsnet is impressive. Sue said that there are now over 100 custom GPTs in operation across the business. “We have a list of them all,” she said. “People can see how they’re being used. They can see other people’s prompts and we’re constantly sharing how people are using them.”

So knowing all this, Mumsnet HQ decided the best response to one of them becoming visible was to try to blame @AmIAComplete ?!! @JustineMumsnet Can you not see how this was a completely inappropriate response?

JaneJeffer · 07/10/2026 15:09

The scale of AI adoption across Mumsnet is impressive. Sue said that there are now over 100 custom GPTs in operation across the business. “We have a list of them all,” she said. “People can see how they’re being used. They can see other people’s prompts and we’re constantly sharing how people are using them.”
Which people? Who are they sharing them with?

Ereshkigalangcleg · 07/10/2026 15:46

JaneJeffer · 07/10/2026 15:09

The scale of AI adoption across Mumsnet is impressive. Sue said that there are now over 100 custom GPTs in operation across the business. “We have a list of them all,” she said. “People can see how they’re being used. They can see other people’s prompts and we’re constantly sharing how people are using them.”
Which people? Who are they sharing them with?

Yes, what does that actually mean? Anyone know?

DrBlackbird · 07/10/2026 16:45

JaneJeffer · 07/10/2026 15:09

The scale of AI adoption across Mumsnet is impressive. Sue said that there are now over 100 custom GPTs in operation across the business. “We have a list of them all,” she said. “People can see how they’re being used. They can see other people’s prompts and we’re constantly sharing how people are using them.”
Which people? Who are they sharing them with?

I assume that she meant MN staff (going by the article) can see other MN staff using the custom GPTs and sharing with each other ways to analyse our data.

Using A.I. for data analytics is what all firms do every day with customer data. Those firms, however, do not have their customers pouring out their souls to them in hundreds of daily posts. And foolishly I haven’t thought of MN as equating to ‘all firms’ because we are not buying a product.

Is the adage, if the product is free, you are the product apt here?

I, for one would appreciate much more information on how those 100 custom GPTs are being used and to what end in order for clarity and transparency.

MrsBennetsPoorNervesAreBack · 07/10/2026 17:12

Ooh, I'm in the guardian!Grin

JaneJeffer · 07/10/2026 17:23

MrsBennetsPoorNervesAreBack · 07/10/2026 17:12

Ooh, I'm in the guardian!Grin

.

Movie Star Hollywood GIF
BiologicalRobot · 07/10/2026 17:54

We introduced title suggestions to help avoid duplicate thread titles

There are only so many human problems and they are bound to be replicated over and over again. Of course you are going to end up with similar titles after twenty odd years. The Relationship forum seems to be in constant groundhog mode at times.

We receive no payment for sending draft text to OpenAI.
So why do it? What do you gain from this if it's not money?

I do NOT consent any drafts i make to be whizzed off to somewhere else before I've pressed the post button. My thoughts and words before I submit them are mine and mine alone. Or so I was given to understand. Unless I'm in some weird, futuristic, matrix type universe already 🤔

ThatLoudCat · 07/10/2026 18:33

Has it been disabled? I just made a thread and there was no AI title suggestion function? Sorry if I've missed that being pointed out already

Ereshkigalangcleg · 07/10/2026 18:35

They’ve suspended it while they work out how to make it opt in and transparent.

randomchap · 07/10/2026 18:39

Ereshkigalangcleg · 07/10/2026 18:35

They’ve suspended it while they work out how to make it opt in and transparent.

Maybe they should have thought about that before putting it live.

They've been dishonest

BlackbirdSingingintheTrees · 07/10/2026 19:26

Is this one (with drafts being sent out, without user knowledge / consent, etc) also a data breach, like this previous instance?

"Mumsnet confirmed to the Guardian that it has now self-referred to the Information Commissioners Office, as it is legally required to do in the event of a data breach."

www.theguardian.com/uk-news/2019/feb/07/mumsnet-reports-itself-to-regulator-over-data-breach

Oliveoy · 07/10/2026 20:24

JustineMumsnet · 06/10/2026 16:29

Hi everyone,
Thanks for continuing to raise questions. I understand why this has unsettled people, particularly the discovery that draft text was being sent to OpenAI while you were writing. I also understand why our incorrect initial explanation has made the subsequent one harder to trust.

Here are some direct answers to questions that have come up on this thread.
We have never tested AI-generated responses to threads on Mumsnet. We do not use AI to write threads or replies.

The draft text sent to OpenAI for a title suggestion is not used to train its models. There is an opt-out from using that data for training, which we obviously chose. We receive no payment for sending draft text to OpenAI.

We introduced title suggestions to help avoid duplicate thread titles and make starting a thread easier. We recognise that removing friction can benefit users, and we approached this as a tool to help people start threads. But I understand that people feel differently about unpublished drafts, which may contain things they ultimately decide not to share. We should have made it clearer how the tool works and given people the option to opt out. As Sue said, the tool is switched off while we look at making it clearer and opt-in.

On moderation, we use AI to help flag potentially problematic posts for our team to review. A human always makes the moderation decision.

Finally, as said previously, we should have investigated more thoroughly before suggesting that something on the OP’s device might be responsible. I’m sorry for that, and for the understandable frustration it caused.

Thank you to the OP for flagging it and to everyone who has taken the time to explain their concerns.

As a business customer of OpenAI you don't choose to opt out of data sharing as you claim you did, you are opted out by default. Only individual users have to opt out.

That's good news for MN users obviously but to claim you actively took steps to protect data when you didn't is a bit disingenuous.

randomchap · 07/10/2026 20:42

There are questions to be answered regarding how mumsnet processes the data.

Notice
Does the user know that their draft, not merely their submitted post, is being sent to OpenAI?

Timing
Is that information presented before the first keystroke is transmitted?

Purpose
Does the notice say that OpenAI receives the draft to generate a suggested title?

Recipient
Does it clearly identify OpenAI rather than hiding it under “third-party providers”?

Scope
Is the entire draft transmitted, or only what is necessary?

Retention
How long does Mumsnet/OpenAI retain the draft?

AI use
Is OpenAI merely processing the text for Mumsnet, or can it use the information for its own purposes?

Special-category data
What happens if someone types medical, sexual, political or other sensitive information into the draft?

Lawful basis
What UK GDPR lawful basis does Mumsnet rely on for this transmission?

International transfers
Where is the data processed and what transfer mechanism applies?

When it comes to expectation, a user reasonably understands that they are writing a post ready to be submitted to Mumsnet, they may not be understanding that the post is being sent to OpenAI in real time.

ICO says that processing should not be unduly unexpected.

Were all of those questions considered before go live?

And that's just regarding using AI to generate titles, the "AIBU Response Schema v1.0:" is still unexplained.

valeriy2 · 07/10/2026 22:28

Two years they were fighting Open Ai now they are in bed with them

www.mumsnet.com/talk/site_stuff/5122770-why-were-taking-legal-action-against-open-ai-and-other-scrapers?page=1&utm_source=chatgpt.com

JaneJeffer · 07/10/2026 22:31

valeriy2 · 07/10/2026 22:28

Still a better love story than Twilight

Opus3 · 07/10/2026 22:53

They use AI for coding then why does everything take so bloody long to fix...

JennyJenny72 · 07/10/2026 23:06

Unbelievable. I found this out from "that" MN reddit sub and now it's in the bloody guardian.

What boggles the mind is that some people are trying to defend MNHQ (and by people, possibly bots). I don't want to talk to bots, I want talk to people.

Yeah yeah, I won't let the door hit me on the way out.

JennyJenny72 · 07/10/2026 23:08

Ereshkigalangcleg · 07/10/2026 18:35

They’ve suspended it while they work out how to make it opt in and transparent.

Yes but only after being busted. It's not defensible IMO.

Ereshkigalangcleg · 07/10/2026 23:14

JennyJenny72 · 07/10/2026 23:08

Yes but only after being busted. It's not defensible IMO.

I’m not defending it.

AstonVillains · 08/10/2026 04:52

Opus3 · 07/10/2026 22:53

They use AI for coding then why does everything take so bloody long to fix...

Because the coding AI is shit?

BlackbirdSingingintheTrees · 08/10/2026 07:01

JustineMumsnet · 06/10/2026 16:29

Hi everyone,
Thanks for continuing to raise questions. I understand why this has unsettled people, particularly the discovery that draft text was being sent to OpenAI while you were writing. I also understand why our incorrect initial explanation has made the subsequent one harder to trust.

Here are some direct answers to questions that have come up on this thread.
We have never tested AI-generated responses to threads on Mumsnet. We do not use AI to write threads or replies.

The draft text sent to OpenAI for a title suggestion is not used to train its models. There is an opt-out from using that data for training, which we obviously chose. We receive no payment for sending draft text to OpenAI.

We introduced title suggestions to help avoid duplicate thread titles and make starting a thread easier. We recognise that removing friction can benefit users, and we approached this as a tool to help people start threads. But I understand that people feel differently about unpublished drafts, which may contain things they ultimately decide not to share. We should have made it clearer how the tool works and given people the option to opt out. As Sue said, the tool is switched off while we look at making it clearer and opt-in.

On moderation, we use AI to help flag potentially problematic posts for our team to review. A human always makes the moderation decision.

Finally, as said previously, we should have investigated more thoroughly before suggesting that something on the OP’s device might be responsible. I’m sorry for that, and for the understandable frustration it caused.

Thank you to the OP for flagging it and to everyone who has taken the time to explain their concerns.

So is this AI gone rogue within Mumsnet like Hugging Face?

Because it seems like the explanations are that either:

A) the AI is doing things beyond what you've instructed it to do, or

B) It's not, this was internally devised, and you are prompting it to produce whole posts / responses

randomchap · 08/10/2026 08:33

JustineMumsnet · 06/10/2026 16:29

Hi everyone,
Thanks for continuing to raise questions. I understand why this has unsettled people, particularly the discovery that draft text was being sent to OpenAI while you were writing. I also understand why our incorrect initial explanation has made the subsequent one harder to trust.

Here are some direct answers to questions that have come up on this thread.
We have never tested AI-generated responses to threads on Mumsnet. We do not use AI to write threads or replies.

The draft text sent to OpenAI for a title suggestion is not used to train its models. There is an opt-out from using that data for training, which we obviously chose. We receive no payment for sending draft text to OpenAI.

We introduced title suggestions to help avoid duplicate thread titles and make starting a thread easier. We recognise that removing friction can benefit users, and we approached this as a tool to help people start threads. But I understand that people feel differently about unpublished drafts, which may contain things they ultimately decide not to share. We should have made it clearer how the tool works and given people the option to opt out. As Sue said, the tool is switched off while we look at making it clearer and opt-in.

On moderation, we use AI to help flag potentially problematic posts for our team to review. A human always makes the moderation decision.

Finally, as said previously, we should have investigated more thoroughly before suggesting that something on the OP’s device might be responsible. I’m sorry for that, and for the understandable frustration it caused.

Thank you to the OP for flagging it and to everyone who has taken the time to explain their concerns.

"We have never tested AI-Generated responses"

Tested? Have you developed something that could do AI responses, just not tested it?

GoodGirlFaith · 08/10/2026 11:50

APlagueOnBothYourHouses · 05/10/2026 06:31

This is crazy! Why would they want AI responses to threads? To bump them if interesting or controversial ones stop getting responses?

I can’t believe the users are being replaced by AI. We’ve got AI posts and then AI responses. We might as well just go and do something else then.

This is interesting and doesn't surprise me one bit as I have suspected it for ages.

I reckon if it is being used then it would be to increase addiction to the site and possibly also to promote certain viewpoints (I think we all know which ones).

I actually commented on a post a few months ago 'AI bot fake post' meaning the I thought the post had been posted by AI bots. I received an email from MN saying my post had been deleted as they didn't like the fact that I was accusing the OP of lying. I wasn't, I was raising my suspicion that the actual post was posted by a bot.

I noticed a couple of years ago that often on a Friday afternoon, a very controversial post would appear on AIBU; usually one of the classic MN subjects that really divides opinion and creates fierce debate. I found myself getting hooked in and feeling compelled to go back to the thread over the weekend. I started to notice increasingly that this would happen on a Friday afternoon; I would notice the OP go up around 3pm; one of the usual controversial subjects. So, I guess it could be a way of bringing people to the site at weekends when MN may usually be quieter as people are with their families.

Swipe left for the next trending thread