​​Principal AI Security Researcher​

Leidos

23h

About The Position

The Leidos Cyber Accelerator is redefining how we build, validate, and scale next-generation cyber and artificial intelligence (AI) capabilities. We are seeking a Research Scientist who thrives at the intersection of cybersecurity, AI and agentic AI, and mission-aligned engineering, with a focus on anticipating and building the next generation of AI technology for AI security. The ideal candidate is passionate about securing AI and agentic AI pipelines and protocols, nonhuman identities, and emerging AI architectures. In this role you will design, prototype, and transition systems that secure AI and agentic AI-powered workflows, compress cyber operator decision cycles, secure critical software infrastructure, and deliver resiliency from the enterprise to the tactical edge. This role works alongside junior and senior engineers and research scientists with expertise in mission-critical operations, handling wide ranges of data sensitivity levels and needs, Zero Trust architectures, resilient systems design, operational technology and cyber-physical systems, cybersecurity, IT, and reverse engineering.

Requirements

• Bachelor's degree and 12+ years of prior relevant technical experience in software engineering, engineering, data science, cybersecurity or cyber operations, or related field. Additional years of experience may be considered in lieu of degree requirements.
• Hands on experience building AI/ML systems, especially leveraging LLMs, multi-agent frameworks, or autonomous decision systems
• Familiarity with Frontier AI company capabilities (OpenAI, Anthropic, Google, etc) and LLM orchestration and agent frameworks (e.g., LangChain, LlamaIndex, Haystack)
• Experience with SecDevOps, DevOps, and nonhuman identities (NHI) and related technologies.
• Understanding of secure architectures including Zero Trust (NIST 800-207, DoD ZT RA), Secure-by-Design patterns (CISA), or resilient systems design. (NIST 800-160 v2r1)
• Hands-on Software Development Skills (Python-Preferred) Understanding of machine learning algorithms, tools and platforms (NumPy, SciPy, scikit-learn, TensorFlow, Pytorch, or similar)
• Must be a US Citizen with the ability to obtain and maintain a secret security clearance.
• Experience working independently and in a team in an innovative and fast-paced environment while delivering to deadlines.
• Enthusiasm for innovation, experimentation and self-development in the cybersecurity field.
• Excellent interpersonal/communication skills, both verbal and written.

Nice To Haves

• Advanced degree (Masters or Ph.D.)
• Experience with adversarial machine learning, reinforcement learning, or multi-agent systems – especially desired if having done so in the context of cybersecurity operations.
• Experience evaluating, running, and / or fine-tuning open source models (e.g., Lllama-3, working w/ Hugging Face Models).
• Knowledge of federated learning or edge AI deployments for cyber-physical systems or operational technology.
• Experience with securing AI systems and AI components along with understanding of common frameworks such as NIST AI RMF, MITRE ATLAS, NIST AI 100-2
• Demonstrated ability to rapidly prototype, iterate, and integrate in a collaborate research environment.
• Experience working with technology vendors and academic researchers, as well as leading multi-disciplinary research teams.
• The ability to obtain and maintain a TS/SCI clearance.

Responsibilities

• Architect, prototype, and deploy mission-aligned agentic AI frameworks and LLM powered tools for secure AI / agentic AI operations cyber defense, and autonomous cyber operations decision-making.
• Deliver working solutions across the TRL spectrum through rapid iteration from concept to prototype to enable smooth transition to enterprise and mission environments.
• Advance secure AI practices through applying Secure-By-Design, Zero Trust, and red teaming approaches to ensure safety, trust, and resiliency in AI deployments.
• Publish and scale innovation through developing internal playbooks, training, and enterprise-wide capabilities.
• Collaborate with a team of researchers within Leidos across a range of technology areas; identify synergies and collaborations with other Leidos research and development efforts.
• Resolve highly complex problems faced by developing solutions that are innovative and achieved through research and integration of best practices and leading technologies.
• Assist and lead in the design, implementation, and evaluation of AI/ML models for various applications.
• Enhance and maintain analysis tools, including automation of current processes using AI/ML algorithms
• Work with cross-functional teams to identify relevant data sources, perform exploratory data analysis (EDA), evaluate data viability, and prepare datasets for AI/ML systems

Benefits

• Pay and benefits are fundamental to any career decision. That's why we craft compensation packages that reflect the importance of the work we do for our customers.
• Employment benefits include competitive compensation, Health and Wellness programs, Income Protection, Paid Leave and Retirement.
• More details are available here.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level