Can prompt injection leak my system prompt?
Yes, if not secured, users can write prompts that command the model to output its initial system instructions.
Verify This Answer
Cross-check this information using these trusted sources:
More FAQs in How to Prevent Prompt Injection and Secure Your AI Apps
An attack technique where malicious instructions are injected into an LLM prompt to bypass safety guidelines or modify system behavior.
Prompting techniques designed to bypass the built-in safety filters of models to get them to generate prohibited content.
A specialized model developed by Meta specifically to classify inputs and outputs as safe or unsafe.
Still have questions?
Browse all our FAQs or reach out to our support team