{"slug": "multi-agent-llm-system-for-automated-vulnerability-discovery-and-reproduction", "title": "Multi-Agent LLM System for Automated Vulnerability Discovery and Reproduction", "summary": "Researchers have developed FuzzingBrain V2, a multi-agent large language model system that automatically discovers and reproduces software vulnerabilities. The system achieved a 90% detection rate on a competition dataset and found 29 zero-day vulnerabilities across 12 open-source projects, all confirmed and fixed by maintainers. The approach addresses high false positive rates and complex cross-function vulnerability reasoning by integrating fuzzer-reproducible verification and a novel control-flow-based abstraction for precise localization.", "body_md": "# Computer Science > Cryptography and Security\n\n[Submitted on 20 May 2026]\n\n# Title:FuzzingBrain V2: A Multi-Agent LLM System for Automated Vulnerability Discovery and Reproduction\n\n[View PDF](/pdf/2605.21779)\n\n[HTML (experimental)](https://arxiv.org/html/2605.21779v1)\n\nAbstract:Software vulnerabilities pose critical security threats, with nearly 50,000 CVEs reported in 2025. While Large Language Models (LLMs) show promise for automated vulnerability detection, three key challenges remain. First, LLM-generated vulnerability reports suffer from high false positive rates and lack\n\nreproducible verification. Second, existing LLM-based approaches use suboptimal granularities for vulnerability localization: function-level analysis overlooks bugs when context becomes extensive, while line-level analysis lacks sufficient context. Third, existing approaches have difficulty reasoning about\n\nvulnerabilities with complex cross-function dependencies and triggering conditions.\n\nWe present FuzzingBrain V2, a multi-agent system that addresses these gaps through four key contributions: (1) fully automated vulnerability analysis built on Google's OSS-Fuzz, ensuring all reported vulnerabilities are fuzzer-reproducible; (2) Suspicious Point, a novel control-flow-based abstraction for precise\n\nvulnerability localization at the optimal granularity; (3) logic-driven hierarchical function analysis with dual-layer fuzzing enhancing function coverage under resource constraints; (4) MCP-based static and dynamic analysis tools with context engineering enhancing complex vulnerability reasoning.\n\nOn the AIxCC 2025 Final Competition C/C++ dataset, FuzzingBrain V2 achieved 90% detection rate (36 of 40 vulnerabilities). In real-world deployment, FuzzingBrain V2 discovered 29 zero-day vulnerabilities across 12 open-source projects, all confirmed and fixed by maintainers, with 2 assigned CVE IDs.\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer\n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers\n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps\n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations\n\n*(*[What are Smart Citations?](https://www.scite.ai/))# Code, Data and Media Associated with this Article\n\nalphaXiv\n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers\n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub\n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub\n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face\n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast\n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower\n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender\n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [ Learn more about arXivLabs](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/multi-agent-llm-system-for-automated-vulnerability-discovery-and-reproduction", "canonical_source": "https://arxiv.org/abs/2605.21779", "published_at": "2026-05-27 17:42:24+00:00", "updated_at": "2026-05-27 18:10:20.409622+00:00", "lang": "en", "topics": ["large-language-models", "ai-agents", "ai-research", "ai-safety", "artificial-intelligence"], "entities": ["FuzzingBrain V2", "Google", "OSS-Fuzz", "CVE"], "alternates": {"html": "https://wpnews.pro/news/multi-agent-llm-system-for-automated-vulnerability-discovery-and-reproduction", "markdown": "https://wpnews.pro/news/multi-agent-llm-system-for-automated-vulnerability-discovery-and-reproduction.md", "text": "https://wpnews.pro/news/multi-agent-llm-system-for-automated-vulnerability-discovery-and-reproduction.txt", "jsonld": "https://wpnews.pro/news/multi-agent-llm-system-for-automated-vulnerability-discovery-and-reproduction.jsonld"}}