Staff AI Security Engineer — Enterprise LLM Security (Remote)

Society for AI Security; United States; Remote; USD 204,000 - 290,000; Full time

Benefits offered by this job

Remote-friendly

Medical coverage for you and dependets

Job summary

affirm is seeking a Staff Security Engineer for Enterprise AI to lead end-to-end security reviews of AI/LLM systems, evaluating architecture, data flows, and tool calls. You will define guardrails, and policy-as-code, and partner across Security, Legal, Privacy, Compliance, IT, and Engineering to scale secure AI adoption.

You will threat model AI/LLM systems, review prompts and MCP configurations, and develop detections to drive risk-based decisions for third-party SaaS vendors such as Notion,

Qualifications

• Experience designing and maintaining security architecture for AI/LLM-based systems.
• Threat modeling AI/LLM apps and securing agentic systems.
• Ability to build security tooling and guardrails, using Python or IaC.
• Familiarity with Kubernetes and AWS; OAuth2, SAML for auth.
• Experience in SOC 2, PCI DSS environments is a plus.

Responsibilities

• Lead enterprise AI security review process across architecture and data flows.
• Threat model AI/LLM systems and drive remediation.
• Review source code, prompts, and tool manifests; craft test cases.
• Design guardrails and policy-as-code for AI security.
• Evaluate third-party SaaS vendors' security postures.
• Lead cross-functional security initiatives and communicate with stakeholders.

Skills

AI security

Threat modeling

Security architecture

Cross-functional leadership

Cloud security

Tools

Python

Terraform

Kubernetes

AWS

CASB

Okta

Job description

affirm is seeking a Staff Security Engineer for Enterprise AI to lead end-to-end security reviews of AI/LLM systems, evaluating architecture, data flows, and tool calls. You will define guardrails, and policy-as-code, and partner across Security, Legal, Privacy, Compliance, IT, and Engineering to scale secure AI adoption.

You will threat model AI/LLM systems, review prompts and MCP configurations, and develop detections to drive risk-based decisions for third-party SaaS vendors such as Notion,