OpenAI Rolls Out 'Advanced' Security Mode for At-Risk ChatGPT and Codex Accounts

image via WIRED

April 30, 2026, 5:30 PM

• •Advanced Account Security requires two physical security keys or passkeys.
• •Account recovery options are limited to recovery keys, backup passkeys, or physical security keys.
• •Users cannot seek help from OpenAI's support team for account recovery.
• •Shorter sign-in windows and session durations are enforced.
• •Users receive alerts for logins and have their ChatGPT conversations excluded from model training by default.
• •Members of OpenAI's Trusted Access for Cyber program must enable Advanced Account Security or use alternative authentication.

OpenAI has launched Advanced Account Security, an optional feature providing enhanced protection for ChatGPT and Codex users. This new security layer, similar to Google's Advanced Protection, aims to make account takeover attacks significantly more difficult. It enforces strict access controls, including the mandatory use of physical security keys or passkeys instead of passwords, eliminating email and SMS for account recovery. This move is part of OpenAI's broader cybersecurity strategy, recognizing the increasing sensitivity of information stored in AI accounts and the need for robust protections.

Read original article