LLMs for Cybersecurity

LLM: Large Language Model

Overview

This field studies how large language models can help identify and explain vulnerabilities, interpret threat intelligence and security logs, and support tasks such as incident response and penetration testing. Common approaches include retrieval-augmented generation, cybersecurity domain adaptation, multi-agent workflows, and integration with static analysis, command-line tools, fuzzers, or execution-based verification. Evaluation spans code datasets, CTI and log benchmarks, and interactive cyber ranges, where results often depend on data quality, tool access, prompts, and scoring methods. Models can assist with specific tasks, but unreliable reasoning, weak generalization to real systems, and the dual-use risks of offensive capability remain central challenges.

Latest papers

CardsList

282 papers23 papers in the last four weeks, down 23% on the four weeks before. 0.2% of all new papers.

  1. Adaptive LLM-Driven Honeypots vs. LLM-Driven Attackers: Does Narrative Coherence Improve Deception?

    Nov 17, 2026Pedro Beltrán-López, Raul Gonzalez Jimenez, Dario Maddaloni +2CybersecurityLLMs for Cybersecurity

  2. LLMs as Validators for Cross-Chain Bridges: Evaluating Source-Level and Transaction-Level Attack Detection on Solana

    Nov 16, 2026Yasir Rahman, Ishraq Samiha, Shahnewaz Karim Sakib +1Software Vulnerability DetectionCybersecurity

  3. ScenarioForge: Resource-Grounded Generation and Execution Validation of Cyber Exercises

    Nov 16, 2026Jaime C. Acosta, Anantaa Kotal, Aritran PiplaiAdversarial Scenario GenerationScenario-Based Testing

  4. TRIAGE-CPG: A Cross-Model Study of Budgeted Static Malware Triage with Large Language Models

    Nov 16, 2026Paul-Andrei Sava, Alexander Küchler, Daniel Kowatsch +4Malware AnalysisLLMs for Cybersecurity

  5. ReconPotBench: Evaluating LLM Honeypots in Multi-Round Linux Reconnaissance

    Nov 16, 2026Qianli Wang, Junchao Wu, Binkai Ou +2LLMs for CybersecurityActive Information Gathering

  6. Condensing Law of Security Information

    Nov 13, 2026Qi Tan, Tianyu Cui, Qi Li +1LLMs for CybersecurityLanguage Model Scaling Laws

  7. I Think, Therefore I Attack: Evaluating Thought Injection Attacks on Reasoning Agents

    Nov 6, 2026Parth Illendula, Garik Livingston, Zain Shamsi +2Prompt Injection Attacks on LLMsPrompt Injection Attacks on AI Agents

  8. MEvince: Knowledge-Constrained LLM Reasoning for Technique Identification from Malware Reports

    Nov 5, 2026Hongyu She, Yiling He, Mengda Xie +4Malware AnalysisLLMs for Cybersecurity

  9. Does Frontier AI Strengthen Cyber Defense? Evidence from Project Glasswing

    Nov 4, 2026Thomas MoskCybersecuritySoftware Security

  10. Analysis of Local Open-Weight LLM Agents on Vulnerability-Reproduction Tasks

    Nov 4, 2026Jorge García-Carrasco, Javier Sanchis, Álvaro Navarro +2LLMs for CybersecurityOpen-Weight Language Models

  11. DoGNAVY-Exploit: Harnessing Parallel Open-Weight Agents for Stage-Aware Vulnerability Exploitation

    Nov 3, 2026Yuxuan Wang, Dadi Guo, Weijie Guo +9Automated Exploit GenerationLLMs for Cybersecurity

  12. StateFence: Task-State Protection Against Evidence-Chain Contamination in Blue-Team LLM Agents

    Nov 2, 2026Haowen Dai, Alejandro Guerra-ManzanaresAI Agent SecurityPrompt Injection Defense

  13. LanG 2.0: Energy-Efficient Agentic AI for Semi-Automated Security Operations at Scale

    Oct 29, 2026Anes Abdennebi, Nadjia Kara, Laaziz Lahlou +1CybersecurityLLMs for Cybersecurity

  14. MetaVul: Agentic Harness Discovery for Vulnerability Detection

    Oct 29, 2026Ira Ceka, Arnav Garg, Ryan Lutz +4LLM Agent HarnessesSoftware Vulnerability Detection

  15. Generation, Not Selection: LLM Proposals for Certified Access-Control Policy Mining

    Oct 27, 2026Gautam Kumar, Ravi Sundaram, Shamik SuralAccess ControlLLMs for Cybersecurity

  16. BUSTA: Teaching Small Language Models to Reason About Binary Execution

    Oct 26, 2026Ritika Poddar, Aidan Fennelly, Deirdre Murphy +3LLMs for Cybersecurity

  17. Repairing Broken Access Control in Web Applications with LLMs: An Empirical Benchmark Study

    Oct 25, 2026Mingrui Liu, Liang He, Xinchu Wang +4Access ControlAutomated Program Repair

  18. From Hype to ROI: An Economic Model for LLM Adoption in Information Systems and Security Operations

    Oct 23, 2026Pellegrino Casoria, Andrea Araldo, Rosario Patanè +1Organizational AI AdoptionLLMs for Cybersecurity

  19. Social-HUMINT: An LLM-Augmented Unified Framework for Measuring Social Media Exposure

    Oct 22, 2026Pellegrino Casoria, Simon Pietro RomanoSocial Media AnalysisCyber Threat Intelligence

  20. SecLog-Bench: A Scalable Benchmark for Cybersecurity Log Reasoning with Scaffolded Evidence

    Oct 22, 2026Yifan Wang, Dongliang Xu, Xiaoyong Li +3CybersecurityLLMs for Cybersecurity

  21. Self-Evolving Static Analysis via LLM Pseudo-Labeling

    Oct 20, 2026Rui Melo, Andre Duarte, Aditya Oke +5Static Code AnalysisSoftware Vulnerability Detection

  22. When Models Know but Can't Linearly Say: Relational Encoding Limits Linear Vulnerability Classification in Code LLMs

    Oct 19, 2026Rui Melo, Andre Catarino, Claudia Mamede +2LLMs for CybersecuritySoftware Vulnerability Detection

  23. HardEval: An Execution-Verified Benchmark for AI-Driven Security Hardening and Compliance Remediation

    Oct 16, 2026Manohar Vellala, Soumya Banerjee, Sandip Roy +2CybersecurityLLMs for Cybersecurity