Guide to checking for unauthorized access on ChatGPT and Claude AI accounts

Identifying Unauthorized Access on AI Platforms

Users of AI platforms like ChatGPT and Claude can determine if their accounts have been compromised by checking active login sessions. This process involves navigating to security or account settings within each platform to review a list of devices currently logged in.

If any unfamiliar devices are listed, users have the option to log out those specific devices or terminate all active sessions to secure their accounts.

ChatGPT Security Steps

For ChatGPT, users access "Active Sessions" through their username, then "Settings," and "Security and Login." From there, they can log out individual unrecognized devices or choose to log out all sessions. To change a password after logging out, users must initiate a password reset through the login page, which sends a six-digit code to their email for verification before setting a new password.

Claude Security Steps

Claude users can find their "Active sessions" under their username, then "Settings," and "Account." Suspicious sessions can be terminated by hovering over them and selecting "Log out" or "Terminate." Claude also provides an option to "Log out of all devices." Claude's login system relies on sending a login link to the user's email address instead of a traditional password.

General Security Recommendations

General security advice for all online services, including AI platforms, includes using unique passwords stored in a password manager and enabling multi-factor authentication (MFA). While ChatGPT offers MFA, Claude does not, instead utilizing email-based login links as its primary authentication method.