OpenAI has announced the introduction of Lockdown Mode for its ChatGPT platform, a feature designed to enhance security against prompt injection attacks, which can potentially lead to data theft and unauthorized control over user information. Lockdown Mode offers a simplified experience, disabling functionalities that could expose sensitive data, such as web browsing, image display, deep research, and file downloads. This mode is particularly relevant for professionals like lawyers and doctors who may input sensitive client or patient information into the chatbot. OpenAI clarifies that Lockdown Mode is not intended for all users but specifically for those who require heightened protection against data exfiltration risks. While the introduction of this feature is a step towards safeguarding user data, it raises concerns about the implications of inputting sensitive information into AI chatbots. Ultimately, the safest approach remains to avoid sharing sensitive data with chatbots altogether.