Skip to content
TechCrunch · AI

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

OpenAI introduced Lockdown Mode for ChatGPT, disabling live web browsing, web image retrieval and display, deep research, and agent mode for self-serve ChatGPT Business accounts and eligible personal accounts.

Why it matters: OpenAI's Lockdown Mode for ChatGPT disables live browsing, image fetching, deep research and agent mode, giving users handling sensitive data a way to limit prompt injection risk.

Read the original ↗Export Markdown