Staff Security Engineer - AI

LinkedIn•Sunnyvale, CA

18d•$152,000 - $248,000•Hybrid

About The Position

As a Staff Security Engineer on the Platform & Product Security team, you will strengthen the security posture of LinkedIn's AI applications, infrastructure, and development workflows — researching emerging threats, leading threat models, designing and implementing secure-by-default frameworks, and building scalable tooling to detect and mitigate vulnerabilities across the company.

Requirements

BA/BS Degree in Computer Science, Cybersecurity, Information Security, or related technical field, or equivalent technical experience.
Hands-on experience or working knowledge of AI/ML or LLM-application security risks (e.g., prompt injection, data poisoning, model/provenance integrity).
5+ years experience in security engineering and in-depth knowledge of application security, authentication and security protocols, cryptography, or supply chain security.
5+ Experience in various security assessment methodologies such as threat modeling, design reviews, penetration testing and vulnerability assessment.
Experience with programming languages such as Java, GoLang or Python.
Experience architecting or implementing tooling that detects and mitigates security vulnerabilities.

Nice To Haves

BS and 8+ years of relevant work experience, MS and 7+ years of relevant work experience, or PhD and 4+ years of relevant work experience.
5+ years experience with programming languages such as Java, GoLang or Python.
Experience securing agentic systems: sandboxing, tool permissioning, and runtime policy enforcement.
Working knowledge of modern AI/ML stack: LLMs and foundation models, RAG architectures, agentic frameworks (e.g. LangChain/LangGraph, CrewAI, AutoGen), and ML pipelines (training, evaluation, inference).
Demonstrated ability to ramp up quickly on AI-specific threats.
Ability to work across teams and communicate concisely and clearly to stakeholders.
Experience with security research, bug bounty or CTF competitions.

Responsibilities

Lead and execute adoption of security controls across AI/ML infrastructure, including model scanning, model signing and provenance, and ML pipeline security (MLOps/MLSecOps)
Architect and build security guardrails and paved roads for agentic workflows: hardening agent SDKs, tool access and permissioning, input/output validation, execution boundaries, sandboxing, and auditability of agent actions
Build automation and red-team/detection tooling that continuously identifies, reproduces, and mitigates AI-specific vulnerabilities at scale; partner with Detection & Response on prompt/tool-call telemetry and AI-incident playbooks.
Lead threat modeling and adversarial red-teaming of AI systems, and maintain a LinkedIn AI threat taxonomy mapped to industry standards. Partner with security and platform teams to embed the resulting threat models and risk assessments into AI platforms, developer tooling, and production services.
Define AI security standards, reference architectures, and policies across model development, data ingestion, training, inference, and agentic systems; drive secure adoption of AI tooling (vendors, copilots, integrations), including data-sharing and sensitive-data-in-prompt boundaries.
Drive cross-functional architecture and engineering decisions across platform engineering, security engineering, ML/AI teams, and product/DevEx to build developer platforms, AI systems, and infrastructure that are scalable, reliable, and secure by design.