245K–285K USD / year
Enforcement Analyst, Agentic Alignment
OperationsHybrid — Boston, MA, San Francisco, CA, Seattle, WA, New York City, NY, Washington, DC
Published on 2026-10-25
These details were extracted automatically from the original listing. They may not be complete or fully up to date — it's worth checking the original job post.
About this role
Investigate harmful or unexpected behavior by AI agents and make evidence-based enforcement decisions under Anthropic’s Usage Policy. The role combines technical analysis of transcripts, tool-use traces, and code with case documentation, escalation, and collaboration to improve detection and policy.
About the company
Anthropic builds reliable and interpretable AI systems, including the Claude family of models, for businesses and developers. The company focuses on AI safety research and practical AI assistants for enterprise and consumer use.
The team
Safeguards, working with Policy, Threat Intelligence, Engineering, Research, and Transparency.
Stack
PythonJavaScriptShell
What you'll do
- Investigate potential agentic misalignment using transcripts, tool-use traces, and code.
- Triage cases by severity and urgency, and escalate through appropriate channels.
- Develop scalable review workflows for agentic cases.
- Make and document enforcement decisions under the Usage Policy.
- Assess cyber-related agent activity and its potential for real-world harm.
- Write evidence-based case reports for technical and non-technical stakeholders.
- Identify trends and feed findings back to policy, detection, and research teams.
- Use AI systems to support investigations and contribute to reports on agentic harm.
What we're looking for
- Read and understand code well enough to determine what happened and why it matters.
- Background in security, cyber threat intelligence, or a related technical field.
- Experience in trust and safety, enforcement, investigations, or a related field.
- Experience conducting structured investigations and making defensible findings.
- Experience supporting formal enforcement actions, including documentation and escalation.
- Clear writing and communication skills for technical and non-technical stakeholders.
- Sound escalation judgment and resilience when working with difficult content.
Nice to have
- Experience using AI systems to accelerate work.
- Comfort working with large datasets, transcripts, logs, and traces.
- Interest in investigating uncategorized AI agent behavior.
- Familiarity with AI safety, alignment, and risks from increasingly autonomous systems.
Benefits
- Generous vacation and parental leave
- Flexible working hours
- Optional equity donation matching