ChatGPT Unusual Activity Detected

Quick Answer

The "unusual activity detected" message means ChatGPT or its security layer thinks your traffic pattern looks risky. This is commonly triggered by VPN use, shared IP addresses, repeated rapid requests, automated browser behavior, or browser extensions that modify traffic.

Symptoms

• A security warning appears before or after login
• Requests are blocked even though your account password is correct
• The issue happens on one network but not another
• Switching to mobile data sometimes removes the warning
• Multiple people behind the same office or campus network see the same problem

Why This Happens

• Your IP address is shared, rotating, or flagged due to VPN or proxy use
• Too many prompts, refreshes, or login attempts happened in a short period
• Browser automation signals or extensions made traffic look abnormal
• A corporate or campus gateway aggregates many users behind one IP
• Security systems are temporarily stricter during service abuse waves

How to Fix

1. Disconnect VPN, proxy, browser relay, or privacy routing features and retry from a normal direct connection.
2. Stop refreshing and repeated sign-in attempts for a while. Aggressive retry behavior can extend the block window.
3. Switch networks. If home Wi‑Fi fails but mobile data works, the original IP reputation is the likely trigger.
4. Restart the router if your ISP gives you a new residential IP after reconnecting. This can help if the current public IP is temporarily flagged.
5. Disable automation-related browser extensions, script tools, and unusual request modifiers.
6. Clear site data and start a clean session after the cooldown period instead of hammering the same blocked state.
7. Use one browser tab only during the retry. Multiple concurrent sessions can look suspicious.
8. Avoid submitting the same prompt over and over during a degraded period, especially if you are already seeing rate or abuse warnings.
9. Test another device on the same network and then another device on a different network to isolate whether the signal is account-based or IP-based.
10. If the block persists on all networks after a long cooldown, contact support with timing and network details rather than continuing blind retries.

If It Still Doesn't Work

• If the message appears together with rate limiting, review chatgpt-too-many-requests
• If disabling VPN fixes it, continue with chatgpt-vpn-not-working
• If login itself fails before the warning clears, go to chatgpt-login-failed