OPENAI has announced a new opt-in account security feature for ChatGPT users at increased risk of targeted hacking attacks, named Advanced Account Security. According to SecurityWeek, the feature focuses on four key areas and is recommended for journalists, researchers, political dissidents, elected officials, and others who handle sensitive tasks.
When enabled, sign-in protection is strengthened by disabling password-based login and requiring physical security keys or passkeys, with OpenAI partnering with Yubico to offer YubiKey devices at a discount. The feature also overhauls account recovery by replacing email- and SMS-based options with backup passkeys, recovery keys, and security keys, though OpenAI notes that its support team will be unable to assist with recovery once these options are enabled.
It shortens sign-in sessions to reduce the risk of takeover and will alert users to logins while allowing management of active sessions. Automatic training exclusion is included, so conversations will not be used to train AI models. The rollout was reported on 4 May 2026.