AI Agent Security & Compliance Lead

Anthropic Limited; San Francisco (CA); Hybrid; USD 255,000 - 345,000; Full time

Benefits offered by this job

Competitive compensation

Equity donation matching

Generous vacation and parental leave

Flexible working hours

Office space in SF

Job summary

Anthropic in San Francisco is seeking a Security Risk & Compliance leader to guide agent governance, risk, and data controls. You will shape policies, collaborate with security, research, and engineering to ensure auditable governance of AI agents, and drive risk decisions with senior stakeholders.

You will define criteria for handling sensitive data, coordinate approvals, and map security controls to international standards.

Qualifications

• 8+ years in security governance, risk, and compliance in tech.
• Experience owning security policies and control standards.
• Ability to design least-privilege access and data controls.

Responsibilities

• Serve as security policy partner for AI agent governance, author policies and standards.
• Define review criteria for access to sensitive data and human approval needs.
• Act as GRC reviewer within engineering approval workflows.
• Define data governance requirements for agents and regulated data.
• Monitor agent behavior and update governance with cross-functional teams.
• Map agent security controls to governance frameworks (ISO 27001, EU AI Act).
• Assess and document agent security risks and drive resolution with engineers.
• Own policy for agent risk acceptances and exceptions.

Skills

Security governance

Policy development

Risk management

Compliance frameworks

Documentation writing

Cross-functional collaboration

Job description

Anthropic in San Francisco is seeking a Security Risk & Compliance leader to guide agent governance, risk, and data controls. You will shape policies, collaborate with security, research, and engineering to ensure auditable governance of AI agents, and drive risk decisions with senior stakeholders.

You will define criteria for handling sensitive data, coordinate approvals, and map security controls to international standards.