OpenAI unveils Lockdown Mode to protect sensitive data from prompt injection attacks

OpenAI has introduced Lockdown Mode, a security feature designed to safeguard sensitive data from malicious prompt injection attacks. This new mode restricts the capabilities of its language models, limiting their ability to access and manipulate potentially sensitive information.

Visit the original source for the full article