Red Team AI Engineer / Analyst

3w3 weeks ago

MUFG

US · Full-time · $98,000 – $123,000

About this role

The Red Team Analyst supports MUFG Red Team operations by executing adversary-emulation activities focused on AI-enabled technologies. This role targets early-career professionals with a computer science background and practical coding experience. They contribute to risk reduction across the enterprise while developing offensive security skills.

Under the direction of senior operators, participate in threat-led assessments, AI security testing, and purple team exercises. Identify how real-world adversaries could exploit application logic, automation pipelines, and AI-assisted capabilities. Execute assigned activities during adversary simulation, assumed-breach scenarios, and control-validation exercises.

Safely conduct testing aligned to approved scopes, deconfliction requirements, and operational guardrails. Support adversarial testing of AI capabilities like prompt injection, misuse scenarios, unintended data access, and guardrail bypasses. Assist in evaluating AI interactions with identity, permissions, APIs, and data sources.

Research emerging attacker techniques targeting modern stacks, CI/CD tooling, and AI/LLM systems. Contribute to Red Team playbooks, test cases, and workflows while supporting engagement reporting. Work four days per week at MUFG offices or client sites and one day remotely.

MUFG invests in talent, technologies, and tools to empower career ownership. Culture emphasizes putting people first, listening to diverse ideas, and collaborating for innovation, speed, and agility. Make a meaningful impact in a global financial group striving for sustainable growth.

Requirements

Bachelor’s degree in Computer Science, Software Engineering, Computer Engineering, or equivalent practical experience
Demonstrated experience through internships, co-ops, academic projects, or personal work
Strong analytical skills
Computer science background
Practical coding experience
Understanding of adversary emulation and offensive security concepts
Familiarity with AI security testing including prompt injection and guardrails
Experience with modern application stacks and developer tooling

Responsibilities

Participate in Red Team missions simulating real-world attacker behavior against applications and platforms
Execute assigned activities during adversary simulation and emulation, assumed-breach scenarios, and purple team exercises
Safely conduct testing activities aligned to approved scopes, deconfliction requirements, and operational guardrails
Support adversarial testing of AI-enabled capabilities including prompt injection, misuse scenarios, unintended data access, and guardrail bypass
Assist in assessing AI-assisted developer tools, copilots, and automation platforms
Research emerging attacker techniques targeting modern application stacks, CI/CD, and AI/LLM systems
Contribute to Red Team playbooks, test cases, and repeatable workflows
Develop and operationalize custom ML/LLM models for cybersecurity use cases

Benefits

Hybrid schedule: four days per week at MUFG office or client sites, one day remote
Investment in talent, technologies, and tools to own your career
Collaborate toward innovation, speed, and agility in a people-first culture
Opportunity to make meaningful impact fostering shared sustainable growth