Note: The job is a remote job and is open to candidates in USA. Teleion is seeking a Senior AI Security Engineer to help clients safeguard AI workloads, agentic systems, and enterprise AI deployments. The role is a hands-on, client-facing position responsible for assessing AI attack surfaces, implementing AI security controls and governance, conducting adversarial testing, and advising clients on AI risk and compliance frameworks.
Responsibilities
- Assess and address the AI and LLM attack surface in production environments: prompt injection (direct and indirect), data leakage through model output, insecure tool use, jailbreaking, and model supply chain risk
- Deploy, configure, and operate AI gateways and LLM proxy layers with centralized control: model allowlists, per-team API keys, rate and budget limits, centralized audit logging, and inline input/output guardrails covering prompt injection, sensitive data egress, credential leakage, and unsafe output
- Design and implement governance frameworks for agentic systems and MCP (Model Context Protocol) tooling: registries, approval workflows, runtime policy enforcement, and identity and authorization patterns for non-human identities
- Implement OAuth-based access, token lifecycle management, and scoped tool permissions for AI agents operating with least-privilege principles
- Extend zero trust controls to AI workloads across identity, device, network, application, and data pillars — with Azure preferred (Entra ID, Defender for Cloud, Key Vault, Purview, Conditional Access)
- Build monitoring and detection for AI systems: prompt behavior analysis, tool invocation patterns, anomalous output detection, and data movement
- Conduct AI red team and adversarial testing using PyRIT, Garak, or equivalent tooling to identify vulnerabilities before adversaries do
- Lead shadow AI discovery programs to identify unsanctioned AI applications and browser extensions across the enterprise
- Perform third-party AI vendor security reviews and assess M365 Copilot security posture including oversharing remediation and sensitivity label enforcement
- Advise clients on AI governance frameworks including OWASP Top 10 for LLM Applications, MITRE ATLAS, NIST AI RMF, and relevant compliance standards
- Perform tasks as assigned
Skills
- Must be living in the following states to be considered: Florida (FL), Georgia (GA) , Illinois (IL), Iowa (IA), Nevada (NV), North Carolina (NC), South Dakota (SD), Texas (TX), Washington (WA), Virginia (VA), Wisconsin (WI)
- 7 or more years of security engineering experience, with at least 2 years directly securing AI, ML, or GenAI systems in production environments
- Demonstrated hands-on experience with LLM and agent attack surfaces: prompt injection, jailbreaking, data extraction, capability chaining, and supply chain risk
- Experience deploying or operating an AI gateway, LLM proxy, or centralized AI control plane with production-grade input/output guardrails and audit logging
- Familiarity with agentic AI risk classes: excessive agency, capability chaining, tool and memory poisoning, and unsafe autonomous action
- Experience implementing identity and authorization patterns for non-human identities in an AI or service-oriented context
- Strong cloud security background with Azure preferred — Entra ID, Defender for Cloud, Key Vault, Purview, and Conditional Access
- AI red team experience using PyRIT, Garak, or equivalent adversarial testing tools
- Proficiency in Python with experience working with LLM APIs (OpenAI, Anthropic, Azure OpenAI) and security evaluation tooling
- Knowledge of OWASP Top 10 for LLM Applications, MITRE ATLAS, and NIST AI RMF
- Azure preferred — Entra ID, Defender for Cloud, Key Vault, Purview, and Conditional Access
- Certifications preferred: CISSP, AZ-500, SC-100, or CCSP
- Familiarity with ISO/IEC 42001 or EU AI Act is a plus
Benefits
- PTO
- Holiday
- 401(k)
Company Overview