A guide to checking and securing accounts on popular AI platforms like ChatGPT, Claude, and Perplexity, emphasizing MFA and password hygiene.
If you suspect unauthorized access to your AI accounts, act quickly—review login activity, reset credentials, and enable multi‑factor authentication to lock out intruders.
Spotting the tell‑tale signs
Most AI platforms log recent sessions; unexpected IP addresses, devices you don’t recognize, or login times that don’t match your schedule are red flags.
Look for sudden changes in usage patterns, such as a spike in token consumption or API calls that you didn’t initiate.
Checking specific platforms
ChatGPT: Open the account settings page, select Security, and review the "Recent activity" table for unfamiliar locations.
Claude: Navigate to the "Account" tab, click Login history, and verify that each entry matches your devices.
Perplexity: The dashboard includes an "API usage" chart; compare the current month’s totals against your known projects.
Immediate remediation steps
- Change the password to a unique, high‑entropy phrase.
- Enable MFA using an authenticator app or hardware key.
- Revoke all active sessions and generate new API keys.
- Update any stored credentials in third‑party integrations.
After resetting credentials, monitor the account for another 48‑hour window to ensure no further anomalies appear.
Best practices to prevent future breaches
Use a password manager to generate and store complex passwords, and never reuse them across services.
Regularly audit third‑party apps with access to your AI accounts and remove any that are no longer needed.
Stay informed about security updates from each platform; they often roll out new protection features.
For a detailed walkthrough, see the TechCrunch guide on detecting compromised AI accounts.