Something new. I think this addon will help alot for many admin.
View attachment 322858
AI Content Moderation for XenForo Enhance your forum's safety with smart, automated content moderation powered by ChatGPT. Key Features: Supports All Compatible ChatGPT Models Select any ChatGPT model that supports the necessary features for...
gnz.is
A bit more about how useful this is - assuming it's the illegal harms version ......
Aside from all the functions, it very much helps to meet criteria for a Child Risk Assessment in addition to a general risk assessment. It enables you to say you have 24/7 moderation for illegal harms, swift take down, pre-moderation (if you tick the box for removal before suspect links are actually posted) as it sends things to the approval queue (if you select that), and "Enhanced moderation". All of which helps class as "Low Risk" via the criteria.
I tested it out with a few things while the site was offline (in a private node) - photos of guns, talk about killing, stroppy posts (eg with an attitude and "you're stupid" on the end.). All were sent for manual moderation and flagged with "weapons" or "bullying and harrassment". Didn't want to test for more serious illegal stuff obviously, but that was enough for me to see it worked well.
I think it's very good. Mine has the list of illegal harms in the list of things to scan for. You can add other things to that (although it might remove more than you want then - haven't tested that fully). Eg if you add Instagram to the list, it'll send any post with an Instagram link for pre-moderation (but might also send a post that mentions Instagram in it and no link).
It's clever that it scans photos and detects a gun eg. In addition to the link scanning and option to scan direct messages.
It's not cheap, but I felt it was worth it. To easily meet a risk assessment and not worry about spam as much.
The other thing that counts as "Enhanced moderation" for a CRA is the Crowd moderation addon (aka user take down).
Two things about enhanced moderation. If you have it, you're more likely to class as low risk for a CRA. If you class as medium risk for a CRA, I think you're required to have enhanced moderation, including hyperlink scanning and user take down.