Stopping alt accounts and ban evaders
Banning someone only works if they stay gone. Alt accounts exist because rejoining is cheap. The fix is layered: make re-entry cost time, signals, and failed checks โ not just a new username.
Understand what you are fighting
Ban evaders reuse devices, browsers, and networks. Some spin up fresh accounts on the same machine minutes later. Others use VPNs and disposable emails to look new. Your verify flow needs to collect signals at join time, not after they already posted.
Require verification before access
If banned users can browse channels before verifying, you already lost. Gate with an unverified role and keep sensitive channels behind the verified role. Every minute they wait is a minute you can run checks.
Use alternative account detection
AuthenticateMe links verification attempts to device fingerprints, browser data, and prior records. When a new account matches someone you already flagged or banned, verification fails under your configured action โ kick, ban, or deny.
Stack entry requirements
Single checks are easy to dodge. Combine what fits your community:
- Minimum account age to kill same-day alts
- Avatar required to filter zero-effort accounts
- Disposable email blocking on the web verify flow
- VPN blocking when evaders hide behind proxies
- MFA requirement for high-trust servers
Log everything
Verification logs give moderators context when someone appeals. You see which check failed, when, and for which account. That beats guessing from a vague "let me back in" DM.
Staff tools for edge cases
False positives happen. Use manual verify for legitimate users caught by aggressive settings. Use reset verification when someone needs to go through again. Use bypass sessions when a trusted member hits a check they should skip once.
Tune over time
Start moderate. Watch logs for a week. Tighten rules if you see patterns. Loosen if real members struggle. Alt defense is ongoing โ AuthenticateMe gives you the knobs; you pick how hard to turn them.
Ready to configure? See all detection features or ask in our support server.