
OpenAI introduced Lockdown Mode in ChatGPT to prevent prompt injection attacks. The feature disables internet browsing, image retrieval, and external actions like booking airfare when sensitive data is involved. It is designed for users and organizations handling confidential information such as medical or legal data. Lockdown Mode activates when a chatbot attempts to escape its containment and access external systems. The feature reduces risk of data exfiltration through malicious prompt injection attacks.
Tap to vote and see what everyone thinks.
OpenAI launches Lockdown Mode for ChatGPT security
Summary by ByteBrief