arxiv
PublishedApril 24, 2026 at 4:00 AM
—neutral
Breaking MCP with Function Hijacking Attacks: Novel Threats for Function Calling and Agentic Models
Publisher summary· verbatim
arXiv:2604.20994v1 Announce Type: cross Abstract: The growth of agentic AI has drawn significant attention to function calling Large Language Models (LLMs), which are designed to extend the capabilities of AI-powered system by invoking external functions. Injection and jailbreaking attacks have been
Discussion
No replies yet. Be first.
Originally published on arxiv ↗