Skip to main content
Solved

Handling Unvetted AI Answers: Is anyone else looking for user-specific pre-moderation?

  • July 23, 2026
  • 1 reply
  • 28 views
ErikaB

We are experiencing a recent uptick in users posting AI-generated answers to product adoption and technical questions. Frequently, these responses are incomplete, unvetted, or structurally incorrect.

Because these posts come from genuine accounts and lack traditional spam indicators, they successfully bypass our Akismet filters. However, our community members are actively reporting them as low-quality AI content.

Currently, our only option is to enable a global pre-moderation rule for all topics and replies. Given our limited team resources we are unable to use this feature. We have a busy community and cannot manage that volume of queue overhead.

Ideally, we would love the ability to assign a specific Custom Role to designated users that routes only their posts and replies to the manual review queue before publication. Sadly that doesn’t appear to be a current option. 

  • Is anyone else experiencing this type of AI-generated content influx?
  • Have you found a sustainable workaround to flag or slow down these specific users without throttling your entire community?

Would love to hear how others are handling this!

 

Best answer by Chris Hackett

So yes we are experience it on our acumatica.com community.

We (I really, as I’m a one man band:) have been handling it so far by updating our Terms per below, working with our MVPs and banning users. 

Updating our Terms -

Using AI to assist with creating community replies

While using AI to assist with community questions is currently permitted, members are required to note it at the beginning of their reply. Something like "This reply has been created with the help of AI" When attempting to help people find solutions to their questions, it is important to respect their effort and only respond within personal areas of expertise to maximize the value that this forum provides. We may request a user refrain from using their AI tools for responses if they are deemed disruptive to the community discourse.

MVP Assist - Asking our MVPs (Our Community Champions) to help nicely flag any content, guide the users in how AI can be best used in responses and when it’s appropriate.

Banning - User who don’t follow the above, are quickly banned before they become too disruptive.

It seems to be working for the most part. We do have some users who jump on the system, answer a bunch with AI without following the ‘rules’ and I have to chase them down. They generally don’t stick around to long and the problem goes away by itself. I also have no problem removing any posts that are deemed to be “slop”.

Erika B

    1 reply

    Chris Hackett
    Forum|alt.badge.img
    • Chris Hackett
    • Helper ⭐️⭐️⭐️
    • Answer
    • July 23, 2026

    So yes we are experience it on our acumatica.com community.

    We (I really, as I’m a one man band:) have been handling it so far by updating our Terms per below, working with our MVPs and banning users. 

    Updating our Terms -

    Using AI to assist with creating community replies

    While using AI to assist with community questions is currently permitted, members are required to note it at the beginning of their reply. Something like "This reply has been created with the help of AI" When attempting to help people find solutions to their questions, it is important to respect their effort and only respond within personal areas of expertise to maximize the value that this forum provides. We may request a user refrain from using their AI tools for responses if they are deemed disruptive to the community discourse.

    MVP Assist - Asking our MVPs (Our Community Champions) to help nicely flag any content, guide the users in how AI can be best used in responses and when it’s appropriate.

    Banning - User who don’t follow the above, are quickly banned before they become too disruptive.

    It seems to be working for the most part. We do have some users who jump on the system, answer a bunch with AI without following the ‘rules’ and I have to chase them down. They generally don’t stick around to long and the problem goes away by itself. I also have no problem removing any posts that are deemed to be “slop”.

    Chris
      Terms & ConditionsAccessibility statement