Optimization to the Rescue: Evading Binary Code Stylometry with Adversarial Use of Code Optimizations

Ben Jacobsen; Sazzadur Rahaman; Saumya Debray

doi:10.1145/3465413.3488574

Optimization to the Rescue: Evading Binary Code Stylometry with Adversarial Use of Code Optimizations

Ben Jacobsen, Sazzadur Rahaman, Saumya Debray

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Recent work suggests that it may be possible to determine the author of a binary program simply by analyzing stylistic features preserved within it. As this poses a threat to the privacy of programmers who wish to distribute their work anonymously, we consider steps that can be taken to mislead such analysis. We begin by exploring the effect of compiler optimizations on the features used for stylistic analysis. Building on these findings, we propose a gray-box attack on a state-of-the-art classifier using compiler optimizations. Finally, we discuss our results, as well as implications for the field of binary stylometry.

Original language	English (US)
Title of host publication	CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021
Publisher	Association for Computing Machinery, Inc
Pages	1-10
Number of pages	10
ISBN (Electronic)	9781450385527
DOIs	https://doi.org/10.1145/3465413.3488574
State	Published - Nov 19 2021
Event	2021 Workshop on Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, CheckMate 2021 - Virtual, Online, Korea, Republic of Duration: Nov 19 2021 → …

Publication series

Name	CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021

Conference

Conference	2021 Workshop on Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, CheckMate 2021
Country/Territory	Korea, Republic of
City	Virtual, Online
Period	11/19/21 → …

Keywords

Adversarial machine learning
Bayesian optimization
Privacy
Stylometry

ASJC Scopus subject areas

Computer Networks and Communications
Computer Science Applications
Software

Access to Document

10.1145/3465413.3488574

Cite this

Jacobsen, B., Rahaman, S., & Debray, S. (2021). Optimization to the Rescue: Evading Binary Code Stylometry with Adversarial Use of Code Optimizations. In CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021 (pp. 1-10). (CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021). Association for Computing Machinery, Inc. https://doi.org/10.1145/3465413.3488574

Optimization to the Rescue : Evading Binary Code Stylometry with Adversarial Use of Code Optimizations. / Jacobsen, Ben; Rahaman, Sazzadur ; Debray, Saumya.

CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021. Association for Computing Machinery, Inc, 2021. p. 1-10 (CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Jacobsen, B, Rahaman, S & Debray, S 2021, Optimization to the Rescue: Evading Binary Code Stylometry with Adversarial Use of Code Optimizations. in CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021. CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021, Association for Computing Machinery, Inc, pp. 1-10, 2021 Workshop on Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, CheckMate 2021, Virtual, Online, Korea, Republic of, 11/19/21. https://doi.org/10.1145/3465413.3488574

Jacobsen B, Rahaman S , Debray S. Optimization to the Rescue: Evading Binary Code Stylometry with Adversarial Use of Code Optimizations. In CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021. Association for Computing Machinery, Inc. 2021. p. 1-10. (CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021). doi: 10.1145/3465413.3488574

Jacobsen, Ben ; Rahaman, Sazzadur ; Debray, Saumya. / Optimization to the Rescue : Evading Binary Code Stylometry with Adversarial Use of Code Optimizations. CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021. Association for Computing Machinery, Inc, 2021. pp. 1-10 (CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021).

@inproceedings{553c8cd2d98e45e8b75e22e2f2b52754,

title = "Optimization to the Rescue: Evading Binary Code Stylometry with Adversarial Use of Code Optimizations",

abstract = "Recent work suggests that it may be possible to determine the author of a binary program simply by analyzing stylistic features preserved within it. As this poses a threat to the privacy of programmers who wish to distribute their work anonymously, we consider steps that can be taken to mislead such analysis. We begin by exploring the effect of compiler optimizations on the features used for stylistic analysis. Building on these findings, we propose a gray-box attack on a state-of-the-art classifier using compiler optimizations. Finally, we discuss our results, as well as implications for the field of binary stylometry. ",

keywords = "Adversarial machine learning, Bayesian optimization, Privacy, Stylometry",

author = "Ben Jacobsen and Sazzadur Rahaman and Saumya Debray",

note = "Funding Information: This research was supported in part by the National Science Foundation under grant no. 1908313. Publisher Copyright: {\textcopyright} 2021 Owner/Author.; 2021 Workshop on Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, CheckMate 2021 ; Conference date: 19-11-2021",

year = "2021",

month = nov,

day = "19",

doi = "10.1145/3465413.3488574",

language = "English (US)",

series = "CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021",

publisher = "Association for Computing Machinery, Inc",

pages = "1--10",

booktitle = "CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021",

}

TY - GEN

T1 - Optimization to the Rescue

T2 - 2021 Workshop on Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, CheckMate 2021

AU - Jacobsen, Ben

AU - Rahaman, Sazzadur

AU - Debray, Saumya

PY - 2021/11/19

Y1 - 2021/11/19

N2 - Recent work suggests that it may be possible to determine the author of a binary program simply by analyzing stylistic features preserved within it. As this poses a threat to the privacy of programmers who wish to distribute their work anonymously, we consider steps that can be taken to mislead such analysis. We begin by exploring the effect of compiler optimizations on the features used for stylistic analysis. Building on these findings, we propose a gray-box attack on a state-of-the-art classifier using compiler optimizations. Finally, we discuss our results, as well as implications for the field of binary stylometry.

AB - Recent work suggests that it may be possible to determine the author of a binary program simply by analyzing stylistic features preserved within it. As this poses a threat to the privacy of programmers who wish to distribute their work anonymously, we consider steps that can be taken to mislead such analysis. We begin by exploring the effect of compiler optimizations on the features used for stylistic analysis. Building on these findings, we propose a gray-box attack on a state-of-the-art classifier using compiler optimizations. Finally, we discuss our results, as well as implications for the field of binary stylometry.

KW - Adversarial machine learning

KW - Bayesian optimization

KW - Privacy

KW - Stylometry

UR - http://www.scopus.com/inward/record.url?scp=85121110077&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85121110077&partnerID=8YFLogxK

U2 - 10.1145/3465413.3488574

DO - 10.1145/3465413.3488574

M3 - Conference contribution

AN - SCOPUS:85121110077

T3 - CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021

SP - 1

EP - 10

BT - CheckMate 2021 - Proceedings of the 2021 Research on Offensive and Defensive Techniques in the Context of Man At The End (MATE) Attacks, co-located with CCS 2021

PB - Association for Computing Machinery, Inc

Y2 - 19 November 2021

ER -

Optimization to the Rescue: Evading Binary Code Stylometry with Adversarial Use of Code Optimizations

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this