OpenAI is expanding Lockdown Mode to all personal ChatGPT accounts. This update brings enhanced security controls to users on free and paid subscription tiers. The rollout aims to reduce risks from external service connections that can expose models to attacks.
Enhanced security controls disable web browsing and agent features for users on free and paid tiers.
Lockdown Mode applies across ChatGPT Free, Go, Plus, Pro, and self-serve Business accounts. It excludes organization SSO-linked accounts using SAML or OIDC protocols. The feature disables several web-connected functions when activated.
The mode limits live web browsing to cached content only. It blocks image support in responses and disables Deep Research capabilities. Agent Mode turns off entirely while Canvas networking faces restrictions. File downloads for data analysis also get blocked under this configuration.
OpenAI introduced Lockdown Mode back in February as a protection layer against potential attacks. The company now confirms the feature does not completely block prompt injections from processed content. Users should treat it as a risk reduction tool rather than a total defense mechanism.
The Active sessions feature joins the rollout across all personal ChatGPT accounts and workspace types. The new Active sessions feature allows users to view and control their active ChatGPT sessions across devices. Organization SSO-linked accounts remain excluded from this functionality alongside Lockdown Mode.
OpenAI is currently rolling out these updates to more ChatGPT users globally. The company has not shared a specific phased schedule or completion date for the deployment. Users will gain access as the rollout progresses across subscription tiers.
Discussion
0 comments
Log in to join the thread with a thoughtful take, question, or correction.