Senior AI Security Researcher

Carnegie Mellon University; False Pass (AK); On-site; USD 150,000 - 190,000; Full time

Job summary

Carnegie Mellon University’s CERT Division at SEI seeks an AI Security Researcher to push the boundaries of AI robustness and cybersecurity. You will collaborate with elite AI and cybersecurity professionals, university faculty, and government representatives to develop new methodologies and publish impactful findings.

The role requires leading research projects, proposing new studies, and presenting recommendations to sponsors and conferences.

Qualifications

• BS in machine learning, cybersecurity, statistics, or related discipline with significant experience
• MS in the same fields with substantial experience
• PhD in the same fields with substantial experience

Responsibilities

• Develop state-of-the-art approaches for analyzing robustness of AI systems
• Apply approaches to understand vulnerabilities in AI systems and attacker tradecraft
• Reverse engineer malicious code to support high-impact customers and develop new analysis methods
• Study and influence AI security and vulnerability disclosure ecosystems
• Evaluate tools and processes developed by industry and AI security research
• Publish results and aid testing and transition of government-funded research

Skills

AI/ML

Cybersecurity

Python

C/C++

Reverse engineering

Numpy

PyTorch

TensorFlow

Education

BS in ML/Cybersecurity/Statistics

MS in ML/Cybersecurity/Statistics

PhD in ML/Cybersecurity/Statistics

Tools

Ghidra

IDA Pro

Job description

Are you a cybersecurity and/or AI researcher who enjoys a challenge? Are you excited about pioneering new research areas that will impact academia, industry, and national security? If so, we want you for our team, where you'll collaborate to deliver high-quality results in the emerging area of AI security.

The CERT Division of the Software Engineering Institute (SEI) is seeking applicants for the AI Security Researcher role. Originally created in response to one of the first computer viruses -- the Morris worm - in 1988, CERT has remained a leader in cybersecurity research, improving the robustness of software systems, and in responding to sophisticated cybersecurity threats. Ensuring the robustness and security of AI systems is the next big challenge on the horizon, and we are seeking life-long learners in the fields of cybersecurity, AI/ML, or related areas, who are willing to cross-train to address AI Security.

As part of the Threat Analysis Directorate, you will join a group of security experts focused on advancing the state of the art in AI security at a national and global scale. Our tasks include vulnerability discovery and assessments for AI systems, evaluation of the effectiveness and robustness of defenses and mitigations for AI systems, reverse engineering AI systems and models, and identifying new areas where security research is needed. We participate in communities of network defenders, software developers and vendors, security researchers, AI practitioners, and policymakers. You'll get a chance to work with elite AI and cybersecurity professionals, university faculty, and government representatives to build new methodologies and technologies that will influence national AI security strategy for decades to come.

You will co-author research proposals, execute studies, and present findings and recommendations to our DoW sponsors, decision makers within government and industry, and at academic conferences. The SEI is a non-profit, federally funded research and development center (FFRDC) at Carnegie Mellon University.

What you'll do:

• + Developstate of the artapproaches for analyzing robustness of AI systems.
• + Apply these approaches to understanding vulnerabilities in AI systems and how attackers adapt their tradecraft to exploit those vulnerabilities.
• + Reverse engineer malicious code in support of high-impact customers, design and develop new analysis methods and tools, work toidentifyand address emerging and complex threats to AIsystemsandeffectivelyparticipatein the broader security community.
• + Study and influence the AI security and vulnerability disclosure ecosystems.
• + Evaluate the effectiveness of tools, techniques and processes developed by industry and the AI security research community.
• + Uncover and shape some of the fundamental assumptions underlying current best practice in AI security.
• + Developthoughtmodels, tools and data sets that can be used to characterize the threats to, and vulnerabilities in, AI systems, and publish those results. You will also use these results to aid in the testing, evaluation and transition of technologies developed by government-funded research programs.
• + Identifyopportunities to apply AI to improve existing cybersecurity research.

Who you are:

• + You have BS in machine learning, cybersecurity, statistics, or related discipline with ten (10) years of experience; OR MS in the same fields with eight (8) years of experience; OR PhD in the same fields with five (5) years of experience.
• + You have a deep interest in AI/ML and cybersecurity with a penchant for intellectual curiosity and a desire to make an impact beyond your organization.
• + You have practical experience with applying cybersecurity knowledge toward vulnerability research, analysis, disclosure, or mitigation.
• + You have experience with advising on a range of security topics based on research and expert opinion.
• + You have familiarity with implementing and applying AI/ML techniques to solving practical problems.
• + You have familiarity with common AI/ML software packages and tools (e.g.,Numpy,Pytorch,Tensorflow, ART).
• + You have knowledge or familiarity with reverse engineering tools (e.g.NSAGhidra, IDA Pro)
• + You have experience with Python, C/C++, or low-level programming.
• + You have experience developing frameworks, methodologies, or assessments to evaluate effectiveness and robustness of technologies.
• + You haveexcellentcommunication skills (oral and written), particularlyregardingtechnical communications with non-experts.
• + You enjoy mentoring and cross-training others and sharing knowledge within the broader community.
• + Candidates withstrong technicalproficiencyin either AI/ML or cybersecurity are welcome to apply, provided ademonstratedintellectual agility and commitmentrequiredfor accelerated learning within the role.

You are able to:

• + Travel to various locations to support the SEI's overall mission. This includes within t