Cybersecurity Engineer, Artificial Intelligence

Wintrust•Rosemont, IL

13h•Hybrid

About The Position

The AI Cybersecurity Engineer will support technical security engineering for Wintrust’s Responsible Artificial Intelligence Alliance (RAIA). This is a hands-on role focused on building robust defense-in-depth for AI systems, specializing in the secure integration of LLM frameworks and the mitigation of adversarial machine learning risks. You will collaborate across multi-disciplined teams to engineer automated security controls, perform red teaming on model deployments, and ensure that AI innovations remain resilient against evolving threat vectors.

Requirements

Bachelor’s degree in Data Science, Computer Science, or AI preferred or equivalent experience
3-5 years of experience in the full Software Development Lifecycle
Specific or relative experience in conducting security reviews of AI-driven applications to identify vulnerabilities in code, model logic, and data integration is a plus, but not necessary
Understanding of machine learning architecture, including the ability to evaluate the security of training pipelines, inference engines, and API-driven workflows from a software engineering perspective

Nice To Haves

Cybersecurity and/or Artificial Intelligence related certifications preferred

Responsibilities

Lead security assessments within the RAIA for Large Language Models (LLMs), Natural Language Models (NLMs), Small Language Models (SLMs), or other projects under the scope of Artificial Intelligence
Participate in RAIA to review use cases based on value, complexity, feasibility, risks, compliance, and alignment with strategic goals
Review AI architectural designs and usage within third-party products for compliance with security best practices, AI security frameworks, and regulatory requirements
Participate in the RAIA to build and maintain standards and frameworks for Wintrust to use to review use cases and help the business develop use cases
Analyze AI-specific security risks, including algorithmic bias, sensitive data exposure, PII leakage, and model poisoning (adversarial attacks), ensuring models remain resilient against manipulation
Demonstrate agility by managing shifting security priorities and “other duties” as AI integration evolves, ensuring Wintrust’s defense posture remains aligned with organizational needs
Experience building and fine-tuning LLMs using frameworks like PyTorch or TensorFlow, with a focus on writing clean, scalable code for data preprocessing, embedding generation, and API integration while maintaining secure coding practices to prevent injection vulnerabilities