AI Account Protection: Why Claude Lacks Two-Factor Authentication

Even the chatbots we use every day to write emails or search for information can become targets of cyberattacks. Accounts on platforms like ChatGPT, Claude, and Perplexity hold conversations, documents, and personal data, making AI account protection a practical skill that every user should know, not just a detail reserved for cybersecurity experts. TechCrunch has published a detailed guide to help users recognize suspicious access and how to respond, platform by platform.

Summary

• Key Points
• Why accounts on ChatGPT, Claude, and Perplexity are targeted by hackers
• How to check active sessions and discover suspicious access
  • ChatGPT
  • Claude
  • Perplexity
• Multi-factor authentication, passwordless login, and resets: the differences that matter
• TechCrunch's guide to AI account protection
• FAQ
  • How can I check if my ChatGPT account has been compromised?
  • Does Claude use passwords for access?
  • Is multi-factor authentication available on all these AI platforms?
  • What should I do if I suspect unauthorized access on Perplexity?

Key Points {#Key_Points}

• Hackers can target accounts on ChatGPT, Claude, and Perplexity just like any other online service.
• Multi-factor authentication is available on ChatGPT and Perplexity, but not on Claude.
• Claude does not use passwords: access is granted via a link sent through email.
• On ChatGPT and Claude, users can see active sessions and disconnect suspicious devices; Perplexity only allows users to log out of all sessions without showing them.
• Password resets on ChatGPT are done with a six-digit code sent via email.

Why accounts on ChatGPT, Claude, and Perplexity are targeted by hackers {#Why_accounts_on_ChatGPT_Claude_and_Perplexity_are_targeted_by_hackers}

A compromised AI account is essentially no different from a hacked social media profile: anyone who gains unauthorized access can read conversation histories, personal data shared with the chatbot, and, in some cases, professional or sensitive information uploaded over time. For this reason, TechCrunch has decided to apply the same approach to AI platforms as it has used for social networks and messaging apps, offering a practical guide to recognize signs of intrusion.

The basic recommendation remains valid for any online service: use unique passwords, store them in a password manager, and enable multi-factor authentication wherever possible. This way, even if someone manages to steal the password, they still cannot access the account without the second verification element.

How to check active sessions and discover suspicious access {#How_to_check_active_sessions_and_discover_suspicious_access}

The first step to determine if an account has been compromised is to check which devices are connected. The three platforms handle this function differently, and this is where the security check for ChatGPT and the other apps differ most distinctly.

ChatGPT {#ChatGPT}

On desktop browsers, simply click on your username in the bottom left, then on "Settings", followed by "Security and Login", and finally "Active Sessions". From here, you can see exactly where the account is connected: if an unrecognized device appears, you can disconnect it individually or click on "Log out all" to close all sessions at once.

Claude {#Claude}

For managing sessions on Claude, the path is similar but goes through "Settings" and then "Account", where you can find the "Active sessions" section. Hovering over an unrecognized session reveals three vertical dots that allow you to click on "Log out" or "Terminate". It is also possible to log out of all devices at once with "Log out of all devices".

Perplexity {#Perplexity}

Perplexity behaves differently from the other two: it does not show at all where the account is connected. Those who fear unauthorized access must go to "All settings" and click on "Sign out of all sessions", confirming the operation. Since there is no visible list of devices, the only way is to log out everywhere and then log back in.

Multi-factor authentication, passwordless login, and reset: the differences that matter {#Multi_factor_authentication_passwordless_login_and_reset_the_differences_that_matter}

The three platforms do not offer the same level of integrated protection, and knowing these differences is essential to understand where more attention is needed. AI multi-factor authentication is available on ChatGPT and Perplexity, while Claude lacks it due to a specific architectural choice by Anthropic.

Instead of asking for a password, Claude always sends a login link via email: it is a passwordless access to Claude that effectively eliminates the possibility of changing credentials, because they simply do not exist. Those who want to log back into their account must enter their email address and wait for the login link.

On ChatGPT, however, the ChatGPT password reset follows a classic procedure: after logging out, you click on "Log in", enter the email, select "Forgot password", and proceed with "Continue". At that point, an email arrives containing a six-digit code to enter on the login page to set a new password. On Perplexity, the mechanism is similar but involves a six-digit code sent at the time of new access, or a direct link contained in the email itself.

This heterogeneity between platforms matters because it shifts the responsibility for security to different points: on Claude, for example, the real entry point becomes the associated email account, not a separate password. Those who entrust their email box to weak credentials risk indirectly exposing access to the chatbot as well.

TechCrunch's Guide to Protecting AI Accounts {#TechCrunchs_Guide_to_Protecting_AI_Accounts}

The guide, authored by Lorenzo Franceschi-Bicchierai, a senior writer at TechCrunch specializing in hacking, cybersecurity, and privacy, provides step-by-step instructions for each of the three platforms, making control accessible even to those without advanced technical skills. The logic is always the same: check active sessions, disconnect anything unrecognized, and, where available, strengthen access with multi-factor authentication.

Anyone suspecting an intrusion should act immediately, disconnecting all devices and changing credentials when possible, especially since a chatbot often retains months of conversations that reveal a lot about its user.

FAQ {#FAQ}

How can I check if my ChatGPT account has been compromised? {#How_can_I_check_if_my_ChatGPT_account_has_been_compromised}

Open ChatGPT in your browser, go to Settings, then Security and Login, and finally Active Sessions to see which devices are connected. From there, you can disconnect a single suspicious device or close all sessions together.

Does Claude use passwords for access? {#Does_Claude_use_passwords_for_access}

No. Claude uses a passwordless login system that sends an access link directly via email.

Is multi-factor authentication available on all these AI platforms? {#Is_multi-factor_authentication_available_on_all_these_AI_platforms}

Multi-factor authentication is available on ChatGPT and Perplexity, but not on Claude.

What should I do if I suspect unauthorized access on Perplexity? {#What_should_I_do_if_I_suspect_unauthorized_access_on_Perplexity}

You need to log out of all sessions through the settings menu and then log back in with your email and the six-digit code received.

Content created with the assistance of artificial intelligence and human editorial review.