OpenAI announced The new feature, he says, will provide additional protection against operative injection attacks, where malicious chatbot instructions are hidden in web pages and other content sources.
Among other things, Lockdown mode will disable live web browsing (you can only access cached content), searching and displaying images from the web (you can still create images), deep exploration, and agent mode.
The company says that even with Lockdown Mode enabled, ChatGPT can still be vulnerable to quick injections, for example, “which may appear in cached web content or an downloaded file and affect the behavior or accuracy of the response.”
But the goal is to reduce the likelihood that sensitive data will be shared in the process.
“It’s lock mode no it’s for everyone,” says OpenAI. “It’s designed for people and organizations who manage sensitive data and want more robust protection against data exfiltration risks associated with operational injection.”
The company says it currently implements a Lockout mode to serve ChatGPT Business accounts as well as eligible personal accounts.
