Enforcement Analyst, Agentic Alignment
Core
Investigate and enforce against harmful and unexpected behavior by autonomous AI agents (agentic misalignment) to ensure safe and beneficial AI systems.
Role type
Senior IC enforcement analyst (agentic alignment)
Builds
Enforcement pipelines and decision frameworks for autonomous AI agents
Domain
AI safety, cyber threat intelligence, and agentic alignment
Required skills
code reading (Python, JavaScript, shell), cyber threat intelligence, trust and safety enforcement, structured investigation, case documentation, technical writing, judgment on escalation
Preferred skills
AI system usage for self-acceleration, large-scale data analysis, curiosity about novel AI agent behaviors, AI safety and alignment knowledge
Technologies
Python, JavaScript, shell scripts, Claude, AI systems
Responsibilities
Review and investigate agentic misalignment via transcripts and tool-use traces; triage cases by severity; operationalize review pipelines; make enforcement decisions under Usage Policy; assess cyber-related agentic activity; write case write-ups; identify patterns and trends; maintain documentation; collaborate with cross-functional teams; use AI to accelerate investigations; contribute to harm trend reports
Seniority
Senior, hands-on IC
