Mumsnet Denies AI Content Creation After Prompt Incident

Mumsnet, a long-established online forum known for discussions on family life, has denied using artificial intelligence to generate content for its message boards. The denial follows an incident where a detailed AI prompt, intended to guide an AI model in writing a typical Mumsnet post, appeared on the site.

The prompt, titled “AIBU Response Schema v.1.0,” was discovered by a user seeking advice on a family matter. The instructions directed an AI to create an “interpersonal reasonableness” post, providing a “clear, forum-style ruling plus brief actionable advice.” The prompt also included definitions for common Mumsnet abbreviations such as AIBU, ND, and DD. It contained directives like “do not diagnose anyone,” “do not demean ND people,” and “prioritize compassion, boundaries, and de-escalation.”

The emergence of the prompt caused a stir among Mumsnet users, with many expressing concern and suspicion. Some described the event as “really creepy” and “so disappointing,” fueling speculation about the site’s use of AI. One user commented, “We might as well just go and do something else then.” Mumsnet, founded in 2000, experienced a 19% drop in monthly unique users in 2025, reaching 6.1 million, and a 600,000 decrease in posts. The company’s latest annual report indicated it was “embedding AI tools to improve productivity and enhance the user experience.”

Research published in Nature suggests that AI assistance on social media can result in “semantic garbage,” with users perceiving AI-generated content as less informative, “robotic,” and “generic.” Following the prompt’s appearance, Mumsnet initially stated it did not use AI for writing threads and that the prompt originated from a tool suggesting thread titles, which utilizes an OpenAI model. This explanation did not satisfy users, some of whom copied the prompt into an AI and found it generated a complete post.

One user, MrsBennetsPoorNervesAreBack, stated, “We are not idiots. Please can you give us a transparent explanation?” Mumsnet then revised its statement, suggesting the prompt did not originate from them and may have transferred from the user’s computer. This further fueled user skepticism, with many believing the prompt must have leaked from Mumsnet’s internal systems.

Justine Roberts, founder of Mumsnet, addressed the situation in a thread, confirming the prompt originated from their system. She explained that the system sends user drafts to OpenAI to suggest titles, but in this instance, it generated a prompt for a full reply. Roberts attributed this to “an error on their side” and reiterated, “We don’t use AI to write threads or replies on Mumsnet.” This explanation also failed to appease users, who felt they were being “fobbed off.”

Users questioned why OpenAI would generate such a specific prompt if it had not been requested by Mumsnet. Roberts clarified to The Guardian that while the prompt’s placement seemed unusual, Mumsnet’s systems do not request, store, or use such prompts. She confirmed Mumsnet has not tested AI tools for drafting posts but does employ AI for coding, advertising operations, conversational analysis, and flagging posts for moderators. Roberts emphasized the importance of Mumsnet’s human-driven content, stating, “Mumsnet works because the people on it are real. Members trust that the advice they get comes from other parents, not a machine.” She concluded that the incident served as a reminder that AI systems can be unpredictable and that users are rightly concerned about Mumsnet remaining a platform written by humans.