Abundant Security Labs
The science that powers Abundant.
We’re operators and researchers who have spent 20+ years building security AI at scale, and who literally wrote the book on it. Here is the published work behind the platform.
Books and Chapters
Not with a Bug, But with a Sticker
Attacks on Machine Learning Systems and What to Do About Them
Ram Shankar Siva Kumar & Hyrum Anderson
Wiley, 2024 — Foreword by Bruce Schneier
Malware Data Science
Attack Detection and Attribution
Joshua Saxe with Hillary Sanders
No Starch Press, 2018
Practical Cryptography in Python
Learning Correct Cryptography by Example
Seth James Nielson & Christopher K. Monson
Apress, 2019
Introduction to Programming for the Independent Student
A Self-Starter's Course on the Principles and Practice of Bending Computers to Your Will
Christopher K. Monson
Independently published, 2020
Cyber Hard Problems
Focused Steps Toward a Resilient Digital Future
National Academies of Sciences, Engineering, and Medicine; Hyrum Anderson (committee member)
Consensus Study Report, 2025
Selected publications
Peer-reviewed and widely-cited work in AI and security from the founding team.
- 18,922 citations
The Llama 3 Herd of Models
Dubey A, Jauhri A, … Saxe J, et al. · arXiv, 2024
- 2,239 citations
The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation
Brundage M, Avin S, … Anderson H, et al. · arXiv, 2018
- 1,168 citations
Deep Neural Network Based Malware Detection Using Two Dimensional Binary Program Features
Saxe J, Berlin K · MALWARE, 2015
- 945 citations
EMBER: An Open Dataset for Training Static PE Malware Machine Learning Models
Anderson HS, Roth P · arXiv, 2018
- 803 citations
Tree of Attacks: Jailbreaking Black-Box LLMs Automatically
Mehrotra A, … Anderson H, et al. · NeurIPS, 2024
- 583 citations
Learning to Evade Static PE Machine Learning Malware Models via Reinforcement Learning
Anderson HS, Kharkar A, Filar B, Evans D, Roth P · arXiv, 2018
- 502 citations
Poisoning Web-Scale Training Datasets Is Practical
Carlini N, Jagielski M, … Anderson H, et al. · IEEE S&P, 2024
- 384 citations
Predicting Domain Generation Algorithms with Long Short-Term Memory Networks
Woodbridge J, Anderson HS, Ahuja A, Grant D · arXiv, 2016
- 364 citations
eXpose: A Character-Level CNN with Embeddings for Detecting Malicious URLs, File Paths and Registry Keys
Saxe J, Berlin K · arXiv, 2017
- 333 citations
Adversarial Machine Learning: A Taxonomy and Terminology of Attacks and Mitigations
Vassilev A, Oprea A, Fordyce A, Anderson H · NIST AI 100-2, 2024
- 313 citations
DeepDGA: Adversarially-Tuned Domain Generation and Detection
Anderson HS, Woodbridge J, Filar B · ACM AISec, 2016
- 211 citations
“Real Attackers Don’t Compute Gradients”: Bridging the Gap Between Adversarial ML Research and Practice
Apruzzese G, Anderson HS, Dambra S, Freeman D, Pierazzi F, Roundy K · IEEE SaTML, 2023
- 194 citations
Purple Llama CyberSecEval: A Secure Coding Benchmark for Language Models
Bhatt M, Chennabasappa S, … Saxe J, et al. · arXiv, 2023
- 164 citations
Parallel PSO Using MapReduce
McNabb AW, Monson CK, Seppi KD · IEEE CEC, 2007
- 163 citations
Malicious Behavior Detection Using Windows Audit Logs
Berlin K, Slater D, Saxe J · ACM AISec, 2015
- 132 citations
The Kalman Swarm: A New Approach to Particle Motion in Swarm Optimization
Monson CK, Seppi KD · GECCO, 2004
Researchers
- Chaired Microsoft's AI Red Team governing board and architected its inaugural red-teaming of production AI systems.
- Co-founded CAMLIS, the Conference on Applied Machine Learning in Information Security.
- Appointed to the 2024 National Academies study on Cyber Hard Problems.
- Advisor to the US and UK governments on AI safety and security.
- 60+ peer-reviewed publications; co-author of Not With a Bug, But With a Sticker.
- Speaker at RSA, BlackHat, and DEFCON. PhD, University of Washington.
- Leads security for Meta's large language models, defending them from application-level attacks.
- Former Chief Scientist at Sophos.
- Principal investigator on multiple DARPA programs at Invincea Labs.
- Led machine-learning security research at Applied Minds.
- Co-author of Malware Data Science; dozens of papers and patents on security AI.
- Speaker at DEFCON, BlackHat, and RSA.
- Senior Security Architect and AI Trust & Safety Lead at Atlassian; formerly Engineering Manager and Senior ML Engineer at Meta.
- CTO at Data Machines Corp., directing DARPA research programs; earlier a Tech Lead at Google.
- Co-author of Practical Cryptography in Python (Apress) and author of Introduction to Programming for the Independent Student.
- Three US patents in intrusive-software and malware detection.
- 20+ peer-reviewed papers in machine learning and optimization (GECCO, CEC, IEEE SSCI, IJCNN).
- Lecturer in cloud-computing security at the Johns Hopkins Information Security Institute. PhD in Computer Science, BYU.
Writing
Our CTO, Josh Saxe, writes regularly on AI, security, and the road to autonomous defense.
Read the Substack