Computing Optimization-Based Prompt Injections Against Closed-Weights Models By Misusing a Fine-Tuning API
Labunets, Andrey, Pandya, Nishit V., Hooda, Ashish, Fu, Xiaohan, Fernandes, Earlence
–arXiv.org Artificial Intelligence
We surface a new threat to closed-weight Large Language Models (LLMs) that enables an attacker to compute optimization-based prompt injections. Specifically, we characterize how an attacker can leverage the loss-like information returned from the remote fine-tuning interface to guide the search for adversarial prompts. The fine-tuning interface is hosted by an LLM vendor and allows developers to fine-tune LLMs for their tasks, thus providing utility, but also exposes enough information for an attacker to compute adversarial prompts. Through an experimental analysis, we characterize the loss-like values returned by the Gemini fine-tuning API and demonstrate that they provide a useful signal for discrete optimization of adversarial prompts using a greedy search algorithm. Using the PurpleLlama prompt injection benchmark, we demonstrate attack success rates between 65% and 82% on Google's Gemini family of LLMs. These attacks exploit the classic utility-security tradeoff - the fine-tuning interface provides a useful feature for developers but also exposes the LLMs to powerful attacks.
arXiv.org Artificial Intelligence
Jan-16-2025
- Country:
- North America > United States (0.67)
- Genre:
- Research Report > New Finding (0.67)
- Industry:
- Government (0.68)
- Health & Medicine > Consumer Health (0.68)
- Information Technology > Security & Privacy (1.00)
- Technology: