AI Security Engineer

About the role

The Information Security organization at Cetera protects employees, advisors, and clients from evolving cyber threats across cloud, SaaS, and emerging AI-enabled technologies. With the expansion of AI capabilities across the enterprise, Cetera is investing in secure AI adoption, governance, and risk mitigation to ensure innovation aligns with regulatory and security expectations.

Responsibilities

• Identify, assess, and mitigate security risks associated with AI/ML systems
• Implement controls aligned to OWASP Top 10 for LLMs and NIST AI RMF
• Perform threat modeling for AI-enabled applications
• Propose and validate technical guardrails to prevent unauthorized AI actions and support secure AI development
• Evaluate third-party AI tools for risk and compliance
• Design and secure cloud-based AI workloads
• Integrate security into CI/CD pipelines
• Partner with IT Risk, Cloud Security, and Engineering teams
• Support incident response for AI and cloud-related threats
• Develop security standards, runbooks, and architecture documentation
• Support audits and regulatory compliance activities

Requirements

5+ years of experience in cloud security, application security, or a combination of both

Hands-on experience securing cloud-native infrastructure and applications

Experience with DevSecOps practices and integrating security into CI/CD pipelines

Understanding of AI/ML concepts and the ability to apply security principles to AI-enabled systems

Familiarity with OWASP (including Top 10 for LLMs) and NIST frameworks (including AI RMF)

Experience in regulated environments (financial services or FINRA preferred)

Strong communication skills across technical and non-technical stakeholders

Preferred Qualifications

• Experience with AWS Bedrock or other AI/ML platforms
• Certifications such as AWS Security Specialty, AWS Certified Generative AI Developer – Professional, CISSP, or CCSP
• Experience with SAST/DAST tools
• Familiarity with AI-specific threats such as prompt injection, model poisoning, and data exfiltration