Principal AI Security Engineer

About the role

ServiceNow is hiring a Principal AI Security Engineer in Petah Tikva for a hands-on architecture role: deep research, secure systems design and prototyping across product lines with minimal direction. You will set platform-wide direction for securing the AI development lifecycle, define the security architecture of agents and models, and represent ServiceNow's AI security position externally while partnering with the AI security research team.

Key responsibilities

• Own the security strategy and architecture for the Agentic AI ecosystem, LLMs and models across product lines.
• Lead threat modeling for complex GenAI and agentic surfaces such as cross-agent autonomy and multi-agent tool orchestration.
• Set the architectural standard for securing model, RAG and data pipelines, including access-control and data-boundary models.
• Define "secure-by-design" for agentic systems and drive its adoption across engineering orgs.
• Represent ServiceNow's AI security posture externally (standards bodies, publications, conferences) and internally to executives.
• Mentor and guide engineers on AI-specific security work.

Required profile

• Bachelor's, Master's or PhD in Computer Science or a related field, or an equally strong track record.
• 10+ years of software engineering experience, including owning the architecture of complex enterprise systems.
• 7+ years in security engineering, with a focus on AI security in recent years.
• Demonstrated engagements in threat modeling and securing LLM, GenAI or agentic systems, ideally with platform-level impact.
• Personal ownership of prompt injection/jailbreak defense, guardrails or AI red-teaming programs.
• Experience designing access-control and data-governance models for model, RAG or data pipelines.

Required skills

• Fluency with the OWASP LLM Top 10, ideally extending or contributing to it.
• Hands-on experience with agentic AI frameworks such as LangChain and LangGraph.
• Java or Python (bonus: 5+ years).
• Bonus: OAuth 2.0, OIDC, PKCE, API Keys, MCP, A2A; open-sourced security tooling; CVEs, publications or invited talks in AI/LLM security; prior AI security strategy at platform level.