Skip to main content
NG-3 is NextGen’s scam detector. It analyzes the images posted in the server and removes the ones it identifies as scams, such as fake screenshots of money withdrawals, cryptocurrency giveaways or fraudulent promo codes.
NG-3 is off by default. You can turn it on from the dashboard, in the Scam detector section.

​
What it does

NG-3 combines several signals from each image, such as its content and the context it is sent in. When it is almost certain that it is a scam, it removes the image and sends an alert.

​
Settings

In Dashboard → Scam detector there are three settings.
Careful: the higher you set the Sensitivity, the greater the risk of false positives. Make sure to keep it in a good balance.
To delete messages, the bot needs the Manage Messages permission in the channels where it is active.

​
How it improves

If NG-3 fails, you can send from the dashboard the scam it missed or the legitimate image it removed by mistake (a false positive). Each submission is reviewed before NG-3 learns from it. The server’s images are analyzed and not stored. The ones you send are kept only until they are reviewed, and then the image is deleted and only a fingerprint that can’t be used to recover it remains. More details in Privacy.

​
Frequently asked questions

Yes, like any detector. If it removes a legitimate image or lets a scam through, you can send it from the dashboard so NG-3 improves.
The second opinion with AI is part of Premium and is not available yet. Everything described on this page is included in the free plan.
More information in How NG-3 works.