Automatically Evading Classifiers: A Case Study on PDF Malware Classifiers

Weilin Xu's talk at
Network and Distributed Systems Symposium 2016
24 February 2016
San Diego, California

For more details see: http://evademl.org/
and the full paper: http://evademl.org/docs/evademl.pdf

Transcript

1. Automatically Evading Classifiers A Case Study on PDF Malware Classifiers Weilin Xu David Evans Yanjun Qi University of Virginia
2. Machine Learning is Solving Our Problems 2 Fake Spam IDS Malware Fake Accounts … …
5. Machine Learning is Eating the World Data Scientist Security Expert 5 ?
6. Machine Learning is Eating the World Data Scientist Security Expert 6 No! Security is different.
7. Goal: Understand classifiers under attack. Results: Vulnerable to automated evasion. Security Tasks are Different: Adversary Adapts 7
8. Building Machine Learning Classifiers 8 Trained Classifier Labelled Training Data ML Algorithm Training (Supervised Learning) Feature Extraction Vectors
9. Assumption: Training Data is Representative 9 Labelled Training Data ML Algorithm Feature Extraction Vectors Deployment Malicious / Benign Operational Data Trained Classifier Training (Supervised Learning)
10. Results: Evaded PDF Malware Classifiers PDFrate* [ACSAC’12] Hidost [NDSS’13] Accuracy 0.9976 0.9996 False Negative Rate 0.0000 0.0056 False Negative Rate with Adversary 1.0000 1.0000 10 * Mimicus [Oakland ’14], an open source reimplementation of PDFrate.
11. Results: Evaded PDF Malware Classifiers PDFrate* [ACSAC’12] Hidost [NDSS’13] Accuracy 0.9976 0.9996 False Negative Rate 0.0000 0.0056 False Negative Rate with Adversary 1.0000 1.0000 11 Very robust against “strongest conceivable mimicry attack”. * Mimicus [Oakland ’14], an open source reimplementation of PDFrate.
12. Variants 12 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach
13. Variants 13 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach /Catalog /Pages 0 /JavaScript eval(‘…’); /Root Modified Parser Extract Me If You Can: Abusing PDF Parsers in Malware Detectors Curtis Carmony,et al.
14. Variants 14 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach /Catalog /Pages 0 /JavaScript eval(‘…’); /Root Mutation Variants From Benign Insert / Replace / Delete
15. Variants 15 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach /Catalog /Pages 0 /JavaScript eval(‘…’); /Root Mutation Variants From Benign 128 546 0 0 Insert / Replace / Delete
16. Variants 16 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach /Catalog /Pages 0 /JavaScript eval(‘…’); /Root Mutation Variants From Benign 128 546 0 0 Insert / Replace / Delete
17. Variants 17 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach /Catalog /Pages 0 /JavaScript eval(‘…’); /Root Mutation Variants From Benign 128 546 0 0 128 0 Insert / Replace / Delete
18. Variants 18 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach /Catalog /Pages 0 /JavaScript eval(‘…’); /Root Mutation Variants From Benign 128 0 Insert / Replace / Delete
19. Variants 19 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach /Catalog /Pages 0 /JavaScript eval(‘…’); /Root Mutation Variants From Benign 128 0 Insert / Replace / Delete
20. Variants 20 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach
21. Variants 21 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach Fitness Function Oracle Target Classifier f(x) Malicious? Score Fitness Score Variants
22. Variants 22 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach Fitness Function Oracle Target Classifier f(x) Malicious? Score Fitness Score Variants Malicious Benign
23. Variants 23 Clone Benign PDFs Malicious PDF Mutation 01011001101 Variants Variants Select Variants ✓ ✓ ✗ ✓ Based on Genetic Programming Automated Evasion Approach
24. Results: Evaded PDFrate 100% 24 Original Malware Seeds
25. Results: Evaded PDFrate 100% 25 Original Malware Seeds Evasive Variants
26. Evaded PDFrate with Adjusted Threshold 26 Original Malware Seeds Evasive Variants Evasive Variants with lower threshold
27. Results: Evaded Hidost 100% 27 Original Malware Seeds
28. Results: Evaded Hidost 100% 28 Original Malware Seeds Evasive Variants
29. 29 Difficulty varies by seed Simple mutations often work Complex mutations sometimes needed. Difficulty varied by targets: PDFrate: 6 days to evade all Hidost: 2 days to evade all Results: Accumulated Evasion Rate
30. Cross-Evasion Effects 30 PDF Malware Seeds Hidost Evasive PDF Malware (against Hidost) Automated Evasion PDFrate 387/500 Evasive (77.4%) 3/500 Evasive (0.6%) Gmail’s classifier is secure?
31. Cross-Evasion Effects 31 PDF Malware Seeds Hidost Evasive PDF Malware (against Hidost) Automated Evasion PDFrate 387/500 Evasive (77.4%) 3/500 Evasive (0.6%) Gmail’s classifier is secure? different.
32. Evading Gmail’s Classifier 32 Evasion rate on : 135/380 (35.5%)
33. Evading Gmail’s Classifier 33 Evasion rate on : 179/380 (47.1%)
34. Conclusion 34 Source Code: http://EvadeML.org Vs. Who will win this arm race?