Secure AI Engineer

The Secure AI Engineer role at Booz Allen Hamilton supports the design, evaluation, hardening, and ongoing improvement of AI-enabled and generative or autonomous AI systems to support safe deployment. The position focuses on security architecture, adversarial testing, and practical mitigations for client solutions.

Responsibilities

• Design secure AI architectures, evaluate AI system resilience, identify vulnerabilities, and implement mitigations that enable safe deployment of generative AI and autonomous AI capabilities
• Advise clients on the secure design, deployment, and governance of AI systems and generative AI applications
• Conduct AI red team assessments targeting large language models (LLMs), retrieval-augmented generation (RAG) systems, AI agents, and multi-agent workflows
• Identify and demonstrate AI-specific attack vectors, including prompt injection, jailbreaks, indirect prompt injection, data poisoning, model inversion, extraction, and other adversarial techniques
• Develop recommendations and technical mitigations to improve AI system security, resilience, and trustworthiness
• Build, test, and evaluate agentic AI systems using modern orchestration frameworks while assessing their security posture
• Design security testing methodologies, automation, and tooling for AI applications throughout the development lifecycle
• Adapt and extend open-source AI frameworks, security tools, and evaluation platforms to meet client-specific requirements
• Stay current with emerging AI threats, vulnerabilities, research, and industry best practices

Requirements

• 5+ years of experience building, deploying, or evaluating AI/ML systems
• 3+ years of experience developing applications using LLMs and modern AI frameworks
• Experience with red teaming AI systems, Adversarial AI, or AI Governance
• Knowledge of AI security principles, secure AI development practices, AI attack techniques, and AI risk management
• Bachelor's degree

Technologies

• Microsoft’s PyRIT
• Garak
• Inspect AI
• OWASP GenAI tools
• LangGraph
• CrewAI
• AutoGen
• Semantic Kernel
• OpenAI Agents SDK
• Vector databases
• Kubernetes
• CI/CD pipelines
• Infrastructure-as-Code

Work Model

• Onsite: work will primarily be performed full-time at a customer facility
• Hybrid: expected to work from a Booz Allen or customer facility, in alignment with leadership expectations and the needs of the role
• Remote: generally expected to have cameras on during meetings; there may still be occasions when in-person work is required at a Booz Allen or customer facility

Location

Washington, DC (onsite)

Compensation

USD 77,600 to 176,000 per year

Benefits

• Health, life, disability, financial, and retirement benefits
• Paid leave
• Professional development
• Tuition assistance
• Work-life programs
• Dependent care
• Recognition awards program acknowledging employees for exceptional performance and superior demonstration of values
• Full-time and part-time employees working at least 20 hours a week on a regular basis are eligible to participate in Booz Allen’s benefit programs

Identity Verification and Interview Requirements

• Complete an identity verification process that leverages advanced biometrics and artificial intelligence to ensure authenticity and protect against identity fraud
• Be on camera during interviews and assessments
• Booz Allen reserves the right to take your picture to verify your identity and prevent fraud

Candidate AI Usage Policy

Use of artificial intelligence (AI) or other tools to assist with responses during interviews (whether in-person or virtual) is prohibited unless permission is explicitly provided.

Commitment to Non-Discrimination

All qualified applicants will receive consideration for employment without regard to disability, status as a protected veteran, or any other status protected by applicable federal, state, local, or international law.