We’ve launched a new AI Antispam system designed to catch smart spam, scams, and inappropriate content that classic filters usually miss.
The module works in the background, scoring messages by risk level and applying actions based on your settings.
💰 Pricing (Pro feature)
To use AI Antispam, your group needs a Rating of at least 4 RP.
Costs are:
• 1 RP / hour to keep AI protection active
• ~4 RP per 200 messages scanned
• 18+ Adult Content Filter is free and can be used independently
⚙️ How to enable it
1. Open Moderation Settings → AI Antispam
2. Turn the main toggle ON
3. Choose which filters you want:
• Spam Detection (PRO) – scans text and captions for spam, scams, suspicious links, bio-ads etc.
• Adult Content Filter (FREE) – blocks 18+ / NSFW in supported media.
• Check comments only – scans only messages that are replies to channel posts in the linked discussion group (messages that are not replies to a channel post are ignored).
4. Select what to scan:
• Text
• Photos
• Videos
• Stickers
• Animations
Tip: if you only care about text spam, you can leave video/image scanning off to reduce processing.
🎚 Sensitivity & actions
AI Antispam assigns each message a confidence score. You decide what happens at different levels:
• Medium suspicion (yellow zone) – messages are reported to admins for review, no automatic punishment.
• High suspicion (red zone) – messages trigger an immediate punishment.
You can configure:
• What the high-risk action is: Delete, Mute, Kick, Ban, etc.
• Optional warning logic, for example: delete the first few messages, then apply a mute.
We recommend starting with the default thresholds and adjusting based on how many false positives/false negatives you see.
🚀 Why it’s useful
AI Antispam works asynchronously and doesn’t slow down your chat. It’s designed to catch:
• Sophisticated spam patterns
• Scam and phishing content
• Tricks like homoglyphs and bio-based advertising
Once configured, it quietly runs in the background and surfaces only the cases that need attention.