OpenAI has introduced a new ChatGPT feature called Lockdown Mode, designed to offer stronger protection against prompt injection attacks.
These attacks happen when harmful instructions are hidden inside webpages, uploaded files, or other content sources that a chatbot may read.
When Lockdown Mode is enabled, several features are restricted to reduce the risk of sensitive data being exposed. Live web browsing is turned off, meaning ChatGPT can only access cached content. The feature also disables the retrieval and display of images from the web, although users can still generate images. Deep research and agent mode are also turned off.
OpenAI says Lockdown Mode does not make ChatGPT completely safe from prompt injection attacks. Malicious instructions could still appear in cached web content or uploaded files, and they may still affect the accuracy or behavior of a response.
The main goal of the feature is to lower the chances of sensitive information being shared during an interaction. OpenAI says the mode is meant for users and organizations that handle sensitive data and want stricter protection against data exfiltration risks linked to prompt injection.
The company says Lockdown Mode is not designed for everyone. It is currently being rolled out to self-serve ChatGPT Business accounts and eligible personal accounts.
If this article helped you, please consider supporting our work. Every small contribution keeps Abijita.com independent and running.
