OpenAI has introduced Lockdown Mode for ChatGPT, a new security feature aimed at reducing the risk of prompt injection attacks and data exfiltration. When enabled, it disables live web browsing, web image retrieval, deep research, and agent mode — limiting attack surfaces while still allowing cached content and image generation. OpenAI acknowledges the mode doesn't fully eliminate prompt injection risks but significantly reduces the chance that sensitive data is exposed. It is targeted at users and organizations handling sensitive data, and is currently rolling out to self-serve ChatGPT Business accounts and eligible personal accounts.
154 Impressions