中文

恶意软件改头换面:通过修改可执行字节突破基于机器学习的静态分析

密码学与安全 2021-10-26 v2 计算机视觉与模式识别

摘要

受深度神经网络(DNNs)在各领域变革性影响的推动,研究人员与杀毒软件厂商提出了直接从原始字节进行恶意软件检测的 DNNs,无需人工特征工程。本工作中,我们提出一种将二进制多样化技术与优化框架交织的攻击,以误导此类 DNNs 同时保持二进制文件的功能。与先前的攻击不同,我们的攻击操纵作为二进制文件功能组成部分的指令,这使其尤其难以防御。我们在白盒与黑盒设定下针对三个 DNNs 评估了我们的攻击,发现其成功率常接近 100%。此外,我们发现我们的攻击可欺骗某些商业杀毒软件,在某些情况下成功率达 85%。我们探索了若干新旧防御手段,并识别出某些可挫败超过 80% 规避尝试的防御。然而,这些防御仍可能易受攻击规避,因此我们主张用不依赖机器学习的方法增强恶意软件检测系统。

关键词

引用

@article{arxiv.1912.09064,
  title  = {Malware Makeover: Breaking ML-based Static Analysis by Modifying Executable Bytes},
  author = {Keane Lucas and Mahmood Sharif and Lujo Bauer and Michael K. Reiter and Saurabh Shintre},
  journal= {arXiv preprint arXiv:1912.09064},
  year   = {2021}
}

备注

Code for transformations at https://github.com/pwwl/enhanced-binary-diversification. Presentation at https://dl.acm.org/doi/10.1145/3433210.3453086. An author of a related work [32] contacted us regarding our characterization of their defense (Sec 2.2). They point out that our attack is not within the stated scope of their defense, but agree their defense would be ineffective against our attack