Papers with third-party system prompts downloaded
PARASITE: Conditional System Prompt Poisoning to Hijack LLMs (2026.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are increasingly deployed via third-party system prompts downloaded from public marketplaces. |
| Approach: | They propose a framework that optimizes system prompts to trigger LLMs to output compromised responses only for specific queries. |
| Outcome: | The proposed framework achieves up to 70% F1 reduction on targeted queries with minimal degradation to general capabilities. |