Truace
Subscribe

Mumsnet denies using AI for content after prompt appears on message board

Mumsnet has denied using AI to produce content after detailed instructions to help an AI model write one of its popular “am I being unreasonable” (AIBU) posts appeared on a Mumsnet message board. Users of the long-established forum, which is known for its frank discussions of family tensions, were vexed when the AI prompt appeared in response to a user seeking advice about a family dilemma involving a neurodivergent uncle and a sick child. The user, called AmIAComplete, had asked for “a sanity check” from other…

By the Truvace desk2 min read1 sourceJournalismTRV-2026-1309· cite & verify
Mumsnet denies using AI for content after prompt appears on message board

In brief

On 7 October 2026, Mumsnet users spotted a detailed AI instruction block appearing in a thread where a user sought advice about a family conflict involving a neurodivergent uncle and a sick child. The block, labeled AIBU Response Schema v.1.0, instructed an AI to produce a forum-style ruling with actionable advice. After initial denials, founder Justine Roberts confirmed it originated from Mumsnet's title-suggestion system that sends drafts to OpenAI, calling it an error on their side.

The episode matters because it shows how even peripheral AI tooling that handles user drafts can leak into public view and undermine trust in peer-support communities built on authentic personal experience. It raises unresolved questions about data handling when drafts are sent to third-party models, transparency of AI use, and whether productivity gains from AI for coding, ads and moderation outweigh perceptions of robotic, generic content.

Main points

  1. Mumsnet denied using AI to write threads after a prompt titled AIBU Response Schema v.1.0 appeared in response to a user seeking advice about a neurodivergent uncle and sick child.
  2. Founder Justine Roberts confirmed the prompt came from its system which sends users' drafts to OpenAI to suggest a title, but said this time it sent a prompt for a whole reply and "looks like an error on their side".
  3. Users described the incident as "really creepy" and "so disappointing" and questioned transparency, with one saying "I just don't know anymore what's real and what isn't sometimes!"

The problem

An AI prompt leak on Mumsnet vexed users who called it creepy and disappointing, fueling speculation about AI-generated threads and concerns that AI assistance creates semantic garbage seen as robotic and generic.

The rundown

The incident began when user AmIAComplete asked for a sanity check about a family dilemma and received a detailed instruction block titled AIBU Response Schema v.1.0 that told an AI to write an interpersonal reasonableness post with definitions for AIBU, ND and DD and rules like do not diagnose anyone and prioritize compassion, boundaries, and de-escalation.

Mumsnet initially said the text came from a title-suggestion tool using an OpenAI model, then suggested it transferred from the user's computer, before founder Justine Roberts confirmed it came from its system that sends drafts to OpenAI and said "We don't use AI to write threads or replies on Mumsnet." Users remained unconvinced, noting the prompt generated a full post when copied into an AI.

Context includes a 19% drop in monthly unique users in 2025 to 6.1 million and 600,000 fewer posts, alongside the company's stated strategy of embedding AI tools, and external research in Nature warning about semantic garbage from AI assistance in social media.

Sources

  1. JournalismThe Guardian2026-10-07

The debate