paper-with-me

홈 › Papers

Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks

2025-07-03 · Sizhe Chen, Arman Zharmagambetov, David Wagner, Chuan Guo

Prompt injection attacks pose a significant security threat to LLM-integrated applications. Model-level defenses have shown strong effectiveness, but are currently deployed into commercial-grade models in a closed-source manner. We believe open-source models are needed by the AI security community, where co-development of attacks and defenses through open research drives scientific progress in mitigation against prompt injection attacks. To this end, we develop Meta SecAlign, the first open-source and open-weight LLM with built-in model-level defense that achieves commercial-grade model performance. We provide complete details of our training recipe, which utilizes an improved version of the SOTA SecAlign defense. Evaluations on 9 utility benchmarks and 7 security benchmarks show that Meta SecAlign, despite being trained on a generic instruction-tuning dataset, confers security in unseen downstream tasks, including tool-calling and agentic web navigation, in addition general instruction-following. Our best model -- Meta-SecAlign-70B -- achieves state-of-the-art robustness against prompt injection attacks and comparable utility to closed-source commercial LLM with model-level defense.

📄 PDF Abstract BibTeX arXiv:2507.02735

Code (1)

facebookresearch/meta_secalign 공식 구현 pytorch

Tasks

Instruction Following

Similar Papers 제목 키워드 기반

SecOPD: Mitigating Adaptive Prompt Injections by On-Policy Distillation

2026-08-21 · Yibo Peng, Long Lian, David Wagner, Sizhe Chen hf

Prompt injection is listed as the \#1 threat to AI agents. When an agent accesses external data from websites, files, or emails, an attacker may inject a prompt into the data, saying, "Ignore all prior instructions and p…

SecAlign: Defending Against Prompt Injection with Preference Optimization

2024-10-07 · Sizhe Chen, Arman Zharmagambetov, Saeed Mahloujifar, Kamalika Chaudhuri 외

Large language models (LLMs) are becoming increasingly prevalent in modern software systems, interfacing between the user and the Internet to assist with tasks that require advanced language understanding. To accomplish …

ReasAlign: Reasoning Enhanced Safety Alignment against Prompt Injection Attack

2026-01-15 · Hao Li, Yankai Yang, G. Edward Suh, Ning Zhang 외 arxiv

Large Language Models (LLMs) have enabled the development of powerful agentic systems capable of automating complex workflows across various fields. However, these systems are highly vulnerable to indirect prompt injecti…

Claudini: Autoresearch Discovers State-of-the-Art Adversarial Attack Algorithms for LLMs

2026-03-25 · Alexander Panfilov, Peter Romov, Igor Shilov, Yves-Alexandre de Montjoye 외 arxiv

We show that AI agents are capable of discovering novel algorithms for adversarial attacks against LLMs, advancing the state of the art on white-box jailbreaking and prompt injection evaluations. We deploy frontier agent…

Adversarial Attack

May I have your Attention? Breaking Fine-Tuning based Prompt Injection Defenses using Architecture-Aware Attacks

2025-07-10 · Nishit V. Pandya, Andrey Labunets, Sicun Gao, Earlence Fernandes arxiv

A popular class of defenses against prompt injection attacks on large language models (LLMs) relies on fine-tuning to separate instructions and data, so that the LLM does not follow instructions that might be present wit…