# שיטות עבודה מומלצות להתמודדות עם משתמשי ספאם ופוסטים של תגובות GPT

**URL:** <https://meta.discourse.org/t/best-practices-dealing-with-spam-users-and-gpt-reply-posts/272525>\
**Category:** Community Building\
**Created:** [22 ביולי,‏ 2023,‏ 10:38pm UTC](https://meta.discourse.org/t/best-practices-dealing-with-spam-users-and-gpt-reply-posts/272525 "2023-07-22T22:38:26Z")\
**Posts on this page:** 1\
**Showing post:** 1

<div class="post-metadata">

**Author:** ![jord8on](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/jord8on/32/166809_2.png) [@jord8on](https://meta.discourse.org/u/jord8on)\
**Post date:** [22 ביולי,‏ 2023,‏ 10:38pm UTC](https://meta.discourse.org/t/best-practices-dealing-with-spam-users-and-gpt-reply-posts/272525/1 "2023-07-22T22:38:26Z")

</div>

**Context** : _We started a community several years ago, right as the pandemic hit strong, then had to devote all resources to other initiatives for several years. We had a lot of participation and then when we stopped community programming and management, the forums slowed to a halt. We’re about to invest more time in implementing our community strategy and management, so this question is pressing the most!_

* * *
  
Right now we are noticing that a lot of the users joining the community right now are either:
1. Joining to post a link back to their spammy affiliate keyword-focused website
2. Joining and posting what very much appears as an LLM / GPT generated reply/post

Here’s an example of a post that hit our moderation queue yesterday:

Note: The user just created their account that same day.

> **Click to view a screenshot of moderated post**
>
> ![Snapshot 2023-07-22 at 16.23.44](https://global.discourse-cdn.com/meta/original/4X/c/b/2/cb2463cc70884caab7dee3e2bef60b4868417687.jpeg)

Our Discourse system message was sent to the user:

> **Click to view a screenshot of canned SYSTEM message**
>
> ![Snapshot 2023-07-22 at 16.22.43](https://global.discourse-cdn.com/meta/original/4X/f/b/4/fb4759ac9be9b1fe721c9ad3c3d9ba941deb26cc.png)

This post from a new user was in response to [this 3-year old post from Chris Anderson](https://forums.pickleballist.com/t/pickleball-forum-introductions-start-here/196/3).

When you read the moderated post from the new user, it’s pretty clear that this message is disingenuous and a response that’s been generated by GPT or something similar.

Are there optimal ways to prevent these types of users from showing up in the community?

Are there ways to prevent users from entering LLM canned responses, for whatever self-serving or malicious reason they may want to be doing this?

Any input would be great!

---

_[View the full topic](https://meta.discourse.org/t/best-practices-dealing-with-spam-users-and-gpt-reply-posts/272525)._