CV

Downloadable PDF available here!

Education

• B.S. in Computer Science, United States Air Force Academy, 2014
• M.S. in Information Security, Carnegie Mellon University, 2016
• M.S. in Machine Learning, Carnegie Mellon University, 2024
• PhD in Electrical and Computer Engineering, Carnegie Mellon University, 2024

Highlighted Experience

• Member of Technical Staff at Anthropic (Frontier Red Team)
  • July 2024 to Present
  • Evaluates, measures, and improves the cyber capabilities of frontier large language models to support their safe and robust deployment. Builds benchmarks and realistic evaluations for autonomous vulnerability discovery and exploit development (e.g., ExploitBench, ExploitGym), assesses model performance in cyber competitions and against real-world software, and develops defensive applications through partnerships with external organizations (e.g., Mozilla, PNNL). Regularly publishes findings on the Anthropic Frontier Red Team blog.
• Student Researcher at Google AI
  • Summer 2022
  • Worked with Dr Nicholas Carlini on the challenges of evaluating randomized machine learning defenses. Pre-print of paper available here.
• Summer Intern at Massachusetts Institute of Technology Lincoln Laboratory
  • Summer 2021
  • Worked with Dr Ross Allen on an intrinsic augmentation to cooperative reinforcement learning that incentivizes accommodative behavior in agents, resulting in demonstrated zero-shot coordination. Paper published at AAMAS 2022 (more details).
• Deep Learning Researcher at Department of Defense (DoD) Research Directorate (RD)
  • Spring 2017 to Fall 2019
  • Customized dense neural nets, recurrent neural nets, generative adversarial networks, and reinforcement learning networks to aid in solving Intelligence Community (IC) problems. Attended NeurIPS and other AI-related conferences.
  • Authored a technical paper presented at internal conferences; the work was pushed into production as a pathfinding use case for deep learning analytics.
• Machine Learning Researcher at Center for Communications Research La Jolla
  • Summer 2018
  • Co-wrote 2 technical papers submitted for publication. Executed 100+ hours of experimentation with top university mathematicians to solve DoD’s hardest problems. Implemented experimental ML process derived from research.
• Technical Project Lead at DoD Joint Artificial Intelligence Center
  • Fall 2018 to Spring 2019
  • Focused on applying ML algorithms for the Humanitarian Assistance and Disaster Relief National Mission Initiative. This is a particularly demanding job that requires constant mission justification to external non-technical organizations and rapid objective pivoting in a highly uncertain environment.

Publications

• Incalmo: An Autonomous LLM-assisted System for Red Teaming Multi-Host Networks

  Brian Singer, Keane Lucas, Lakshmi Adiga, Meghna Jain, Lujo Bauer, and Vyas Sekar. Incalmo: An autonomous LLM-assisted system for red teaming multi-host networks. In Proceedings of the 47th IEEE Symposium on Security and Privacy, May 2026.

• Training Robust ML-based Raw-Binary Malware Detectors in Hours, not Months

  Keane Lucas, Weiran Lin, Lujo Bauer, Michael K. Reiter, Mahmood Sharif. Training Robust ML-based Raw-Binary Malware Detectors in Hours, not Months. In Proc. CCS '24.

• Group-based Robustness: A General Framework for Customized Robustness in the Real World

  Weiran Lin, Keane Lucas, Neo Eyal, Lujo Bauer, Michael K. Reiter, Mahmood Sharif. Group-based Robustness: A General Framework for Customized Robustness in the Real World. In Proc. NDSS '24.

• RS-Del: Edit Distance Robustness Certificates for Sequence Classifiers via Randomized Deletion

  Zhuoqun Huang, Neil G. Marchant, Keane Lucas, Lujo Bauer, Olga Ohrimenko, Benjamin I. P. Rubinstein. RS-Del: Edit Distance Robustness Certificates for Sequence Classifiers via Randomized Deletion. In Proc. NeurIPS '23.

• Adversarial Training for Raw-Binary Malware Classifiers

  Keane Lucas, Samruddhi Pai, Weiran Lin, Lujo Bauer, Michael K. Reiter, Mahmood Sharif. Adversarial Training for Raw-Binary Malware Classifiers. In Proc. USENIX Security '23.

• Perspectives from a Comprehensive Evaluation of Reconstruction-based Anomaly Detection in Industrial Control Systems

  Clement Fung, Shreya Srinarasi, Keane Lucas, Hay Bryan Phee, Lujo Bauer. Perspectives from a Comprehensive Evaluation of Reconstruction-based Anomaly Detection in Industrial Control Systems. In Proc. ESORICS'22.

• Constrained Gradient Descent: A Powerful and Principled Evasion Attack Against Neural Networks

  Weiran Lin, Keane Lucas, Lujo Bauer, Michael K. Reiter, Mahmood Sharif. Constrained Gradient Descent: A Powerful and Principled Evasion Attack Against Neural Networks. In Proc. ICML'22.

• Any-Play: an Intrinsic Augmentation for Zero-Shot Coordination

  Keane Lucas, Ross E. Allen. Any-Play: an Intrinsic Augmentation for Zero-Shot Coordination. AAMAS'22.

• Malware Makeover: Breaking ML-based Static Analysis by Modifying Executable Bytes

  Keane Lucas, Mahmood Sharif, Lujo Bauer, Michael K. Reiter, Saurabh Shintre. Malware Makeover: Breaking ML-based Static Analysis by Modifying Executable Bytes. In Proc. AsiaCCS'21.

• Creating xBD: A Dataset for Assessing Building Damage from Satellite Imagery

  Ritwik Gupta, Bryce Goodman, Nirav Patel, Ricky Hosfelt, Sandra Sajeev, Eric Heim, Jigar Doshi, Keane Lucas, Howie Choset, Matthew Gaston. Creating xBD: A Dataset for Assessing Building Damage from Satellite Imagery. In Proc. CVPR Workshops, '19.

Presentations

• Workshop Presentation: Adversarial Training with a Surrogate

  Keane Lucas, Alec Jasen and Lujo Bauer. Adversarial Training with a Surrogate. NeurIPS NewInML Workshop. December 2020

• Conference Presentation: Malware Makeover: Breaking ML-based Static Analysis by Modifying Executable Bytes

  Keane Lucas, Mahmood Sharif, Lujo Bauer, Michael K. Reiter, Saurabh Shintre. Malware Makeover: Breaking ML-based Static Analysis by Modifying Executable Bytes. In Proc. AsiaCCS'21.

• Conference Presentation: Any-Play: an Intrinsic Augmentation for Zero-Shot Coordination

  Keane Lucas, Ross E. Allen. Any-Play: an Intrinsic Augmentation for Zero-Shot Coordination. AAMAS'22.

• Conference Presentation: Adversarial Training for Raw-Binary Malware Classifiers

  Keane Lucas, Samruddhi Pai, Weiran Lin, Lujo Bauer, Michael K. Reiter, Mahmood Sharif. Adversarial Training for Raw-Binary Malware Classifiers. In Proc. USENIX Security '23.

• Claude: Climbing a CTF Scoreboard Near You

Projects

• Doom Bot: deep reinforcement learning agent that plays DOOM from raw visual input (GIF-filled presentation)

Top Achievements

• 2019 National Defense Science and Engineering Graduate (NDSEG) Fellowship Awardee (166 awarded from 2928)
• Joint Commendation Medal – awarded August 2019
• 2018 DoD Research Directorate Military Performer of the Year, awarded Research Directorate and Math Research coins
• 2018 JASON Presenter – Conducted hour-long technical DoD AI presentation for the elite government advisory group
• 2016 Carnegie Mellon University Information Networking Institute Leadership Annual Award
• Undergraduate Cyber Training Top Graduate, Honor Graduate, and Major General Robert E. Sadler Award
• Honored by Alaska State Legislature (my home state) for ‘exemplary service with dignity and honor’ (2014)
• USAFA Instructor Letter (rare award given upon instructor agreement that this cadet should come back to teach)
• Honor Lists (7 x Dean’s and 4 x Commandant’s) / Dean’s: academic honor / Commandant’s: military honor
• 2010 Graduation High School Valedictorian in class size of 290

Skills

• Data Analysis/Security
  • Introduction to Statistical Data Mining, Graduate Artificial Intelligence, Introduction to Machine Learning, Machine Learning Large Datasets, Search Engines, Cryptography, Secure Software Systems, Network Security
• Systems
  • Software Reverse Engineering, Distributed Systems, Embedded Systems, Languages and Machines
• Languages
  • Python, Java, LaTeX, C, Go, ARM, x86, Octave, Perl, C#, Objective C, SQL

Service and Leadership

• Deputy Flight Commander, 7th Intelligence Squadron, US Air Force
  • Spring 2017 to Fall 2019
  • Led a Flight consisting of 70+ airmen across 4+ directorates in research, capabilities, and IT support
• President, Graduate Organization Information Networking Institute, Carnegie Mellon University
  • Fall 2015 to Spring 2016
  • Led graduate organization comprising ~200 students from ~15 countries and served as a dept strategic board member
• Commander, US Air Force Academy Cyber Defense Exercise
  • Spring 2014
  • Led an interdisciplinary team of 40 cadets to operate a secure network for the NSA Cyber Defense Competition

Other Experience

• Rapid Capability Scripter, Air Force Red Team (57th Information Aggressor Squadron), US Air Force
  • Winter 2017/2018
  • Delivered 28 new capabilities to simulate adversary behavior for large scale war games (RED FLAG)
• Cyber Security Intern, NASA Jet Propulsion Laboratory
  • Summer 2015
  • Found security bugs in Key System and built bare-metal to VM process for pen-testing and private cloud transition