ChatGPT Advanced Account Security feature released for high-risk users
Summary
Hide ▲
Show ▼
OpenAI introduced an opt-in Advanced Account Security feature for ChatGPT and Codex users, particularly targeting high-risk groups such as journalists, researchers, political dissidents, and elected officials. The feature replaces password-based authentication with physical security keys or passkeys, eliminates SMS and email-based account recovery in favor of backup passkeys, and shortens sign-in session durations to reduce account takeover risk. Enabled users are excluded from AI training datasets and receive enhanced login alerts and session management capabilities.
Timeline
-
04.05.2026 12:29 1 articles · 1h ago
OpenAI launches Advanced Account Security for ChatGPT and Codex accounts
OpenAI released an opt-in Advanced Account Security feature for ChatGPT and Codex users, requiring security keys or passkeys for sign-in and replacing SMS/email recovery with backup passkeys and recovery keys. The feature shortens sign-in sessions, provides login alerts, enables session management, and excludes conversations from AI training. Support for account recovery is disabled once enabled.
Show sources
- OpenAI Rolls Out Advanced Security for ChatGPT Accounts — www.securityweek.com — 04.05.2026 12:29
Information Snippets
-
The Advanced Account Security feature disables password-based login for ChatGPT and Codex users who opt in, mandating the use of physical security keys or passkeys.
First reported: 04.05.2026 12:291 source, 1 articleShow sources
- OpenAI Rolls Out Advanced Security for ChatGPT Accounts — www.securityweek.com — 04.05.2026 12:29
-
OpenAI has partnered with Yubico to offer discounted YubiKey devices to eligible users of the new security feature.
First reported: 04.05.2026 12:291 source, 1 articleShow sources
- OpenAI Rolls Out Advanced Security for ChatGPT Accounts — www.securityweek.com — 04.05.2026 12:29
-
Account recovery for users enabling this feature is handled exclusively through backup passkeys, recovery keys, or security keys, with OpenAI support unable to assist once enabled.
First reported: 04.05.2026 12:291 source, 1 articleShow sources
- OpenAI Rolls Out Advanced Security for ChatGPT Accounts — www.securityweek.com — 04.05.2026 12:29
-
Sign-in sessions are shortened when Advanced Account Security is activated, reducing exposure to session hijacking or device compromise.
First reported: 04.05.2026 12:291 source, 1 articleShow sources
- OpenAI Rolls Out Advanced Security for ChatGPT Accounts — www.securityweek.com — 04.05.2026 12:29
-
Users opting into the feature receive real-time alerts for logins and can manage active sessions, while their conversations are excluded from AI model training.
First reported: 04.05.2026 12:291 source, 1 articleShow sources
- OpenAI Rolls Out Advanced Security for ChatGPT Accounts — www.securityweek.com — 04.05.2026 12:29