Updating Community Guidelines to handle AI-generated posts

A much needed discussion as we transition into this world ​@jvdc 

The line between AI-assisted and AI-generated is blurry and I'm not sure we want to police it too heavily. How I’m thinking is more around the intent. Is it adding value? Or is it purely activity farming?

A couple of things that have helped us:

• Context over content: we look at the full picture of an account (post history, profile, engagement pattern) before taking action on any post.
• Community norms over rules: Rather than over-policing, I believe we are trying to set a culture where members know what good contribution looks like. The T&C addition was part of that. A community guidelines doc is also an additional touch point that will come out soon. I prefer having this disclaimer loud and clear & would probably expand it more over time. 😅

The bigger challenge is activity farming - posting AI content just to climb the leaderboard.

Haven't seen a heavy load of that yet here, but would love to hear how others are handling it!

We reached out to and notified users that their answers look to be AI generated. Some indicated their want/desire to communicate in their native language but ran their post through AI to translate it which resulted in the post coming off as generated. We worked with those users directly to help get spaces created for them and helped with wording on replies.

We immediately banned users, especially those with public email domains, who started to reply to handfuls of accounts all at once with AI replies. Most went back and tried to change their replies after the post was approved to include what Ill call links.

My Ts and Cs https://community.jamf.com/p/get-started We didn’t directly call out how to use AI but might in the future.

In the beginning of 2026 we’ve started to discuss how to deal with AI generated content in our community. In this process we also included our super users to give their point of view and their preferences.

We quickly established that it doesn’t make sense to fully ban AI content, as this would be impossible to enforce. Also, there are AI use cases we found to be useful and helpful to users and the community as a whole. Such as grammar checks or optimization the format of longer texts. 

We used those discussions as a basis to write down AI Guidelines for our community members.

You can find it here (in German): KI-Richtlinien - Unsere Regeln für Nutzung und Kennzeichnung von KI-Inhalten in der O₂ Community

In short:

• AI-generated content is not permitted
  • Content generated solely by AI will be removed
• We support the use of AI as a tool, but we do not accept AI as the creator of community content, because the O₂ Community is and will always be a platform by people, for people
• The exchange of personal impressions, individual opinions, and original creations is the core (and the purpose) of the O₂ Community
• AI-optimized content is permitted, but must always be clearly labeled as such
  • For content that has been optimized using AI but not (properly) labeled, we reserve the right to label it retroactively or remove it after consulting with the authors

Then we go on and substantiate the different use cases, what is permitted, what is not, and - very important - how to properly and transparently label if AI has been used, and how.