{"ID":2890625,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.19598","arxiv_id":"2507.19598","title":"MOCHA: Are Code Language Models Robust Against Multi-Turn Malicious Coding Prompts?","abstract":"Recent advancements in Large Language Models (LLMs) have significantly enhanced their code generation capabilities. However, their robustness against adversarial misuse, particularly through multi-turn malicious coding prompts, remains underexplored. In this work, we introduce code decomposition attacks, where a malicious coding task is broken down into a series of seemingly benign subtasks across multiple conversational turns to evade safety filters. To facilitate systematic evaluation, we introduce \\benchmarkname{}, a large-scale benchmark designed to evaluate the robustness of code LLMs against both single-turn and multi-turn malicious prompts. Empirical results across open- and closed-source models reveal persistent vulnerabilities, especially under multi-turn scenarios. Fine-tuning on MOCHA improves rejection rates while preserving coding ability, and importantly, enhances robustness on external adversarial datasets with up to 32.4% increase in rejection rates without any additional supervision.","short_abstract":"Recent advancements in Large Language Models (LLMs) have significantly enhanced their code generation capabilities. However, their robustness against adversarial misuse, particularly through multi-turn malicious coding prompts, remains underexplored. In this work, we introduce code decomposition attacks, where a malici...","url_abs":"https://arxiv.org/abs/2507.19598","url_pdf":"https://arxiv.org/pdf/2507.19598v1","authors":"[\"Muntasir Wahed\",\"Xiaona Zhou\",\"Kiet A. Nguyen\",\"Tianjiao Yu\",\"Nirav Diwan\",\"Gang Wang\",\"Dilek Hakkani-Tür\",\"Ismini Lourentzou\"]","published":"2025-07-25T18:11:10Z","proceeding":"cs.CL","tasks":"[\"cs.CL\",\"cs.AI\",\"cs.CR\",\"cs.LG\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}