What Is a Discord AI Bot?
What actually makes a bot 'AI-powered,' and how Airwavy's two separate AI systems — AI Chatbot and AI Moderation — differ.
Quick answer
An AI Discord bot uses a language model to generate responses or understand content dynamically, instead of only running fixed, pre-programmed commands. Airwavy has two separate AI systems: AI Chatbot, a conversational assistant with memory and image understanding, and AI Moderation, a 14-category safety classifier that only ever alerts staff or deletes a message — never bans, kicks, or times out anyone.
What you'll learn
- What actually makes a bot 'AI-powered' versus a bot with static commands
- Airwavy's two separate AI systems, and what each one is actually for
- How the AI Chatbot holds a conversation, reads images, and — on Premium — searches the web
- How AI Moderation acts as a second-opinion safety layer, and its real limits
- What's free versus Premium for each system
Why "AI bot" means different things
"AI bot" gets used loosely. Some bots just have an AI-sounding name attached to the same fixed command list every other bot has. A genuinely AI-powered bot uses a language model to generate a response or understand content that wasn't written in advance — which changes what you can actually expect from it: can it hold a conversation, understand an image, or look something up live, or is it just picking from a list of canned replies?
What Discord itself offers
Discord doesn't ship a built-in AI conversation feature for servers. Any AI chat experience a server has comes from a bot the server owner chose to add — Discord provides the platform (slash commands, message content, permissions) that a bot like Airwavy builds an AI feature on top of.
Airwavy's two AI systems
These are genuinely separate systems with separate jobs. Conflating them is the most common source of confusion.
AI Chatbot — a conversational assistant
Responds when a message starts with an Airwavy mention, or to any message in a configured channel. Holds short-term memory (about ten exchanges per member per server, expiring after 30 minutes of inactivity), reads attached images, and offers preset personalities (Airwavy, chill, comedian, storyteller, zen, tech, fitness). Server Premium and Personal Premium add live web search, on by default, for current information like news or prices — capped and clearly labeled as a search-backed answer.
AI Moderation — a safety classification layer
Not a chat feature. Reviews message candidates that deterministic signals already flagged across 14 safety categories, and only ever alerts staff or deletes a confirmed-unsafe message — it never bans, kicks, or times out anyone. It's Server Premium and opt-in, layered on top of Airwavy's regular moderation tools rather than replacing them.
Where to start
- Want a conversational assistant members can talk to? Enable AI Chatbot from the dashboard — see the dedicated setup guide for the exact steps.
- Want an extra safety net for ambiguous moderation cases? Enable AI Moderation from the Moderation feature and choose which of the 14 categories apply to your community.
- The two are independent — enable one, both, or neither, depending on what your server actually needs.
Common mistakes
- Assuming AI Moderation will ban or kick automatically — it never does; it only alerts staff or deletes a confirmed-unsafe message.
- Enabling AI Chatbot without publishing any rules for how members should use it — set expectations before turning it on server-wide.
- Treating chatbot answers as authoritative — AI output can be inaccurate; it's assistance, not a source of truth for high-stakes questions.
- Assuming a custom chatbot personality can loosen its safety behaviour — core safety instructions apply regardless of the configured style.
Troubleshooting
I enabled AI Moderation and nothing seems to happen
That's often correct — it only reviews candidates that deterministic signals already flagged as suspicious, and only ever alerts or deletes. It's a second opinion on edge cases, not a general content scanner.
The chatbot gave a wrong or outdated answer
AI output can be inaccurate, and free-tier answers come from the model's own knowledge only. Server Premium and Personal Premium add live web search for current information.
I'm not sure which system flagged something
AI Moderation actions are distinct from AI Chatbot replies in the dashboard and any configured moderation log — check the alert channel configured for AI Moderation specifically.