Model Serving Security

API security, rate limiting, and inference endpoint protection

Newest first · 2 reviewed on this page

paper2024USENIX Security 2024Reviewed

Rodrigo Pedro, Daniel Castro, Paolo Molina +1

Demonstrates how prompt injection can be chained with traditional web attacks (SQL injection, XSS) in LLM-integrated applications.

paper2024arXiv preprintReviewed

Sagiv Antebi, Noam Azulay, Edan Habler +3

Analyzes security risks of custom GPTs in the OpenAI GPT Store including prompt leakage, data exfiltration, and malicious GPTs.