ChatGPT Lockdown Mode Rolls Out to All Personal Accounts

OpenAI is making its Lockdown Mode security feature available to a much wider audience. After initially launching the feature in February for a limited group of highly security-conscious users, the company has now started rolling it out to all personal ChatGPT accounts, including Free, Go, Plus, and Pro subscribers.

The feature is also becoming available for self-serve ChatGPT Business accounts, giving more users access to enhanced protections designed to reduce potential security risks while using AI-powered tools.

Users will be able to enable Lockdown Mode through ChatGPT Settings under the Security section once the feature reaches their account.

What Lockdown Mode Does

Lockdown Mode is designed for users who prioritize security over convenience. When enabled, ChatGPT restricts or disables several features that rely on web access or external services in order to reduce potential attack surfaces.

One of the biggest changes affects web browsing. ChatGPT can only access cached content while Lockdown Mode is active, which means search results may be limited, unavailable, or outdated. The AI may also stop displaying images in standard responses or retrieving images from the web.

Users can still upload image files directly, and image generation remains available in regions and plans where the feature is normally supported.

Several Advanced Features Become Unavailable

OpenAI’s security-focused mode also disables several newer ChatGPT capabilities. Deep Research and Agent Mode are unavailable, Canvas-generated code cannot access the network, ChatGPT cannot download files for analysis, and web-connected workflows become restricted.

However, manually uploaded files continue to work normally, allowing users to analyze documents and data that they provide directly.

OpenAI notes that Lockdown Mode is not a complete defense against prompt injection attacks. Malicious instructions may still appear inside uploaded documents or cached web content. The company says the feature focuses on reducing opportunities for sensitive information to leave a conversation rather than eliminating all attack vectors.

Active Sessions Feature Gets Wider Rollout

Alongside the Lockdown Mode expansion, OpenAI has also announced broader availability for Active Sessions.

The feature allows users to review every device and browser currently signed into their account. From a centralized dashboard, users can terminate active sessions on other devices if they notice suspicious activity.

Active Sessions provides information such as device or browser details, application context, approximate location, sign-in date and time, trusted device status, and identification of the current session.

According to OpenAI, Active Sessions will not be available for accounts connected through organization-managed single sign-on systems such as SAML or OIDC.

The expanded rollout reflects OpenAI’s ongoing efforts to give users greater control over account security and privacy. The company has recently introduced major improvements to ChatGPT’s memory system and has continued expanding Codex with new job-specific plugins and developer-focused capabilities.