Beyond Passwords: OpenAI Boosts ChatGPT Protection

OpenAI is enhancing user protection with its new optional Advanced Account Security feature for ChatGPT and Codex. This initiative aims to significantly deter account takeover attacks by implementing stricter access controls. The company highlighted that AI services are increasingly handling sensitive personal and professional information, making robust security essential.

This advanced security tier mandates the use of physical security keys or passkeys, replacing conventional passwords. Furthermore, traditional account recovery methods like email and SMS have been eliminated, with users now relying on recovery keys, backup passkeys, or physical security keys. OpenAI has partnered with Yubico to provide security key bundles to users opting for this enhanced protection.

A critical aspect of Advanced Account Security is that support teams will no longer have access to recovery options, thereby preventing social engineering attacks targeting customer service. Users will also experience shorter sign-in windows and receive alerts for new logins. Additionally, conversations will be opted out of model training by default for these secure accounts.