Agent Security Lead

The Opportunity

‍

Autonomous agents don't stop to ask permission — they act, and that means every action needs a guardrail already in place before it's taken. As our Agent Security Lead, you'll design and enforce the boundaries that keep our agents safe, predictable, and aligned with policy, even when they're making thousands of decisions a minute.

‍

This isn't a compliance-checkbox role. You'll be embedded with the agent runtime team, shaping how permissions, escalation paths, and kill-switches are architected from the ground up.

‍

What a Typical Engagement Looks Like

‍

An agent is granted access to a new third-party API. Before it ships, you're the one who defines its permission scope, sets rate and action limits, and builds the audit trail that lets us reconstruct exactly what it did and why — after the fact if something goes wrong.

‍

You'll also run red-team exercises against our own agents, probing for prompt injection, scope creep, and unintended chained actions, then work with engineering to close the gaps you find.

‍

Your Responsibilities

‍

• Define and enforce permission boundaries and escalation policies for autonomous agents
• Build audit and traceability systems for every agent-initiated action
• Run adversarial testing against agent behavior to surface security gaps
• Partner with legal and compliance to keep guardrails aligned with evolving AI regulation