Security Engineer 2, AI

Oscar Health•New York, NY

1d•$163,944 - $215,176•Hybrid

About The Position

Oscar is the first health insurance company built around a full stack technology platform and a relentless focus on serving our members. We started Oscar in 2012 to create the kind of health insurance company we would want for ourselves—one that behaves like a doctor in the family. As Senior AI Security Engineer, you'll be at the forefront of safeguarding our digital ecosystem and shaping the secure deployment of artificial intelligence at Oscar. You will collaborate with dynamic teams across Data, Engineering, Privacy, HR, and Compliance to architect cutting-edge security solutions and proactive guardrails that drive safe AI innovation, secure enterprise model adoption, and protect our business from emerging adversarial threats.

Requirements

4+ years of relevant experience in technical security engineering, application security, cloud security infrastructure, or software engineering with a distinct focus on data security or AI frameworks.
Bachelor’s degree in a technical discipline (Computer Science, Data Science, Cybersecurity, or equivalent technical space).

Nice To Haves

Proven experience evaluating LLM security postures, implementing alignment guardrails, and managing cloud security infrastructure (AWS/GCP IAM boundaries).
Comfort and fluency with cutting-edge frontier AI technologies (e.g., Claude Code, Codex) and a strong aptitude for the continuous, rapid adoption of emerging AI engineering tools.
Experience communicating technical AI risk requirements and architectural implications clearly to non-technical business partners, legal, and compliance teams.
Ability to commit, review, and manage code independently across multiple enterprise repositories.
Familiarity with health insurance data regulations (e.g., HIPAA) and their intersections with automated data usage and generative model architectures.

Responsibilities

Lead the development and implementation of a comprehensive AI security framework encompassing secure model adoption, training data pipelines, retrieval-augmented generation (RAG) architectures, and fine-tuning environments to ensure robust safety and compliance.
Architect, deploy, and operationalize automated controls to mitigate risks such as prompt injection, model poisoning, data exfiltration, and insecure output handling while enforcing least-privilege principles across LLMs and agentic workflows.
Lead proactive AI red-teaming exercises and map adversarial tactics against industry frameworks like MITRE ATLAS and the OWASP Top 10 for LLMs to continuously harden systems against modern vulnerabilities.
Develop innovative metrics and reporting frameworks to evaluate the impact and organizational adoption of AI safety policies, providing clear, actionable insights for executives and ensuring strict accountability to compliance and legal stakeholders.
Partner cross-functionally with application and data engineering teams to establish secure-by-default blueprints, helping employees responsibly harness AI technologies while establishing a strong security-first mindset.
Other duties as assigned.