AI Security Research & Red Team Engineer

The role

This role sits on the Red Team inside the company's security emulation and threat detection group. It pairs offensive research into AI and agentic systems with full-scope simulated attacks against the company's infrastructure and products. The engineer works alongside incident response, detection, product reliability, and compliance teams to turn findings into defensive gains.

What you would do

• Track new threats and dig into AI-specific and conventional weaknesses across the company's products and services.
• Probe agentic systems and large language model integrations to map AI attack surfaces and shape security requirements and guidance.
• Run end-to-end adversary emulation against global infrastructure, internal networks, and the product stack.
• Build a measurable method for judging how reliably web application firewall, endpoint detection, and log analytics catch known attacker techniques.
• Work side by side with defenders to convert attack findings into quick detection and response upgrades.
• Mentor colleagues as a technical leader, encouraging hands-on curiosity, ethical hacking, and collaboration.
• Translate intricate exploit findings into risk-focused briefings so executives can direct engineering effort to the highest-impact fixes.
• Run unannounced exercises with the incident response team to sharpen playbooks, on-call processes, and forensic tooling.
• Help the detection engineering group close coverage gaps, author new detection logic, and validate tuning against emerging techniques.
• Feed red team results to product and reliability engineers so systems and processes become harder to exploit.
• Supply empirical evidence to governance and compliance staff to support automated control validation in place of manual audits.

What they asked for

• Four or more years of experience in offensive security, application security, or a closely related field.
• Working knowledge of AI systems, coding assistants, LLMs, prompt design, and agent architectures, including attacks such as prompt injection and jailbreaks.
• Hands-on skill in manual pentesting, writing exploits, or securing cloud environments on AWS, GCP, or bare metal.
• Practical experience using MITRE ATT&CK to chart detection coverage and spot gaps in defensive telemetry.
• Ability to explain a sophisticated zero-day to non-specialists while earning the trust of senior engineers.
• Familiarity with breach and attack simulation platforms and bespoke tooling for payload staging and command-and-control infrastructure.