AI Security Engineer

Periscope Search Group; Toronto; Hybrid; CAD 110,000 - 170,000; Full time

Job summary

Periscope Search Group is partnering with a global design and manufacturing client in Toronto to hire an AI security specialist. You will design secure architectures for AI agents and LLMs across enterprise systems, implementing guardrails and identity controls to prevent data leakage.

You will govern AI usage policies and oversee subscriptions with major providers. The role requires hands-on experience building AI security tooling, leading POCs, and collaborating with cloud security architects

Qualifications

• Experience designing secure AI agents and guardrails in enterprise environments.
• Strong background in cloud security architecture and secure development
• Familiarity with CGP/ITAR/EAR or equivalent regulatory frameworks
• Ability to lead security POCs and governance for AI platforms

Responsibilities

• Design how AI agents and LLMs are secured across enterprise finance, ERP, and aerospace/defense environments.
• Implement non-human identities and access guardrails to prevent data leaks.
• Govern AI usage policy and govern enterprise subscriptions with major providers.
• Lead proofs-of-concept on new AI security tooling and integration with CI/CD pipelines.

Skills

AI security

Cloud security

LLM governance

Security architecture

Regulatory compliance

Tools

Anthropic

OpenAI

Google Cloud

Job description

Location: Toronto, Ontario

Full time

Hybrid up to 3x days / week

Must qualify for CGP or ITAR / EAR

Our client is a global design and manufacturing company with 40+ sites in 13 countries, serving aerospace and defense, healthtech, communications and hyperscale cloud customers. They don't just use AI, they build programs around next-generation AI technology, and their security architecture team is adding a dedicated AI security specialist.

In this role, you'll design how AI agents and LLMs are secured across enterprise finance, ERP and aerospace and defense environments: non-human identities that can reach one data source but not the next, guardrails that prevent accidental disclosure, and secure build pipelines for agents. You'll also govern enterprise subscriptions with Anthropic, OpenAI and Google, set AI usage policy, and lead POCs on new AI security tooling.

This suits someone who has built and deployed AI agents, not just reviewed them. If you know how to build it, you know how to break it, and that is how you'll secure it. You'll join a team of cloud and application security architects and report to the Director of Security Architecture.