OpenAI Rolls Out Advanced Account Security with Hardware Key Anti-Phishing Protection

OpenAI has recently announced the launch of Advanced Account Security (AAS), an optional protection measure for ChatGPT users. Designed for high-risk individuals, it can be voluntarily activated by any user. AAS replaces traditional password verification with software passkeys or hardware security keys, and adds enhanced protections such as login alerts, shortened session durations, and backup recovery keys.

As a key part of this new initiative, digital security company Yubico announced a partnership with OpenAI to release two co-branded hardware security key products: YubiKey C NFC and YubiKey C Nano. Users can authenticate by inserting the key into a device or tapping via NFC, effectively defending against phishing attacks and account hijacking. However, if a user loses their hardware key without having saved a backup recovery key in advance, OpenAI will not be able to help restore account access.

This security upgrade comes as AI chat tools increasingly handle sensitive conversations and business data. By introducing hardware-level authentication, OpenAI aims to significantly reduce the risk of unauthorized access to ChatGPT accounts worldwide. For users who handle confidential information or frequently use AI agents, this move sets a new industry benchmark for account security and signals that AI service providers are accelerating the shift from a "password era" to a "hardware authentication era."