AI Red Team Lead Engineer

The AI Red Team Lead Engineer leads the execution and evolution of offensive security activities focused on AI/ML systems, platforms, and integrations, in addition to traditional enterprise attack surfaces. Owns the design, execution, and reporting of AI-focused red team operations, adversarial testing, and threat emulation exercises targeting models, data pipelines, AI-enabled applications, and supporting infrastructure.

Acts as a senior technical authority and program lead for AI red teaming, partnering with teams to identify, validate, and communicate AI-related risks. Drives maturity through repeatable testing methodologies, automation, custom tooling, and clear articulation of business impact.

- Lead AI Red Team operations, including adversarial testing of:

- Foundation and custom models (LLMs, vision, speech, decision systems)

- Model deployment environments (APIs, plugins, agents, RAG pipelines)

- Design and execute AI-specific threat emulation aligned to real-world adversaries, misuse scenarios, and emerging attack techniques (e.g., prompt injection, data poisoning, model inversion, jailbreaks, supply chain risks).

- Develop and maintain custom AI red team tooling, frameworks, and automation to scale testing and improve repeatability.

- Perform security research into emerging AI attack techniques, model vulnerabilities, and defensive gaps.

- Partner with detection, engineering, and governance teams to support purple-team and control validation activities.

- Contribute to AI security standards, testing guidance, and program strategy.

- Mentor and provide technical leadership to red team engineers.

- Thorough understanding of the applicable information security systems, policies, and procedures

- Effective communication, presentation skills, leadership, problem-solving and analytical skills

- Hands-on experience testing or securing AI/ML systems, including LLMs or other model classes

- Knowledge of AI threat models and attack techniques including, but not limited to, prompt injection, model extraction, training data poisoning, inference abuse, hallucination exploitation

- Familiarity with AI platforms and tooling (e.g., model APIs, orchestration frameworks, evaluation pipelines)

- Significant red team experience, including adversary emulation and multi-stage attack chains

- Proven skill developing proof-of-concept exploits and custom offensive tooling

- Strong understanding of red team and offensive AI techniques and tooling

- Expertise defeating or bypassing endpoint and AI-adjacent security controls (EDR/XDR, API protections, guardrails)

- Experience with cloud, containerized, and AI-hosting environments

- Proficiency in one or more languages (e.g., Python, PowerShell, Go, C/C++, Shell)

This role requires working from a U.S. Bank location three (3) or more days per week.

The salary range reflects figures based on the primary location, which is listed first. The actual range for the role may differ based on the location of the role. In addition to salary, U.S.

Bank offers a comprehensive benefits package, including incentive and recognition programs, equity stock purchase 401(k) contribution and pension (all benefits are subject to eligibility requirements). Pay Range: $126,820.00 - $149,200.00