How to Check if Your AI Accounts on ChatGPT, Claude, and Perplexity Are Hacked
Popular artificial intelligence platforms such as ChatGPT, Claude, and Perplexity can be targeted and breached by unauthorized users, similar to any other online service. According to a security guide published by TechCrunch, users can verify whether unfamiliar devices are accessing their accounts and take concrete steps to terminate active sessions.
What Happened
As AI platforms gain widespread use, unauthorized access has emerged as a risk across major services. While ChatGPT, Claude, and Perplexity handle credentials differently, each provides mechanisms to help users inspect active logins or invalidate current sessions to regain control.
Standard recommendations include keeping unique passwords saved inside a password manager and enabling multi-factor authentication (MFA). While ChatGPT and Perplexity support MFA, Claude does not utilize traditional passwords, relying instead on login links sent directly to a user’s registered email address.
Key Highlights
- ChatGPT: Users can navigate to the bottom-left username menu, choose “Settings,” then “Security and Login,” and open “Active Sessions.” Unrecognized devices can be logged out individually, or users can select “Log out all.” To reset credentials, users must sign out, select “Forgot password” on the login page, and enter a six-digit verification code sent to their email.
- Claude: Under the bottom-left username menu, users can click “Settings” and then “Account” to review active sessions. Sessions can be ended individually by hovering over an entry, clicking the three vertical dots, and choosing “Log out” or “Terminate.” Alternatively, users can choose “Log out of all devices.” Because Anthropic does not use passwords for Claude, relogging occurs via email links.
- Perplexity: Unlike ChatGPT and Claude, Perplexity does not display a list of active devices. Users concerned about potential unauthorized access must click their username in the bottom-left corner, select “All settings,” and choose “Sign out of all sessions.” Users then log back in using an email address and a six-digit verification code or direct email sign-in link.
Why This Matters
Without checking active sessions or invalidating unauthorized logins, third parties could maintain undetected access to an account. Understanding how authentication operates across different AI platforms helps users choose appropriate defense steps, such as resetting passwords on services that support them or relying on email-based authentication links and session terminations on platforms that do not.
What to Watch Next
Users who suspect unauthorized access should sign out of existing sessions immediately across their devices and update their credentials or email protections as supported by each platform.
Frequently Asked Questions
Does Claude support multi-factor authentication?
No. Claude does not offer multi-factor authentication or traditional password logins. Instead, Anthropic sends a login link directly to the account holder’s email address.
Can you view a list of active logged-in devices on Perplexity?
No. Perplexity does not show where an account is currently logged in. To ensure unapproved devices are disconnected, users must use the option to sign out of all sessions simultaneously.
How do you update your password on ChatGPT?
To change a ChatGPT password, log out of the account, select “Log in,” provide the associated email, choose “Forgot password,” and enter the six-digit verification code delivered via email to set a new password.
Source: TechCrunch