New ChatGPT Lockdown Mode Limits Tools That Could Enable Data Exfiltration

OpenAI launches Lockdown Mode for ChatGPT to protect against prompt injection-based data exfiltration attacks.

• •Restricts outbound network requests by disabling web browsing, image support, Deep research, and Agent mode features
• •Intended for users and organizations handling sensitive data that require enhanced security guarantees
• •Available across Free, Go, Plus, Pro plans and ChatGPT Business with optional activation
• •Cannot be used simultaneously with Developer Mode; both modes are mutually exclusive

This summary was automatically generated by AI based on the original article and may not be fully accurate.