Cybersecurity & Adversarial Testing Specialist
Posted 72d ago
Posted by Karokan
karokan.ai
€120–190/hr
0 of 3 steps completed · 0%
Submit your resume or equivalent profile to start. We review background, communication quality, and evidence of relevant execution experience before moving forward.
ⓘ Application steps are reused across roles — you never upload your resume or take the same interview twice.
Role Overview
Stress LLM agents against prompt injection, exfiltration, privilege escalation, and tool abuse in enterprise deployment contexts.
Karokan is working with AI labs and deployment teams across Europe to contract high-judgment specialists who can improve model quality through reliable human input. This role is designed for experts who can translate nuance into structured outputs, work with clear evaluation rules, and contribute directly to the performance of production AI systems.
You will operate in a fully remote setting, collaborate asynchronously, and help shape datasets, reviews, and decision layers that support the next generation of AI products. The work is intentionally rigorous: accuracy, consistency, and written reasoning matter more than generic throughput.
Key Responsibilities
- Design LLM-specific attack scenarios for enterprise copilot environments
- Probe for prompt injection, data exfiltration, and privilege escalation
- Deliver written red-team report with severity ratings and mitigations
Ideal Qualifications
- Background in offensive security, penetration testing, or security research
- Experience with LLM threat modeling or AI security evaluations