Safeguards Account Access & Enforcements Specialist

CS Recruiting; San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Job summary

Anthropic safeguards needs an Operations Specialist to run human review for access to its models and features and to respond when accounts are compromised or misused. You will document decisions against policy and communicate outcomes to customers and internal partners.

The role demands precision, sound judgment, and comfort applying evolving rules, including exposure to sensitive content. Collaboration with policy, enforcement, engineering, and sales is essential.

Qualifications

• 2+ years in trust & safety, fraud, risk, compliance, or another operations role where you made decisions against written policy.
• Ability to hold many detailed, evolving rules in your head and apply them consistently.
• Write clear, concise decision rationales and customer-facing communications.
• Keep meticulous records and understand that every decision needs an audit trail.
• Know when a case is clear-cut and when it needs escalation, and elevate with a crisp summary.
• Learn quickly and adapt as new processes are stood up and revised.
• Work well both independently and with policy, enforcement, engineering, and sales partners.
• Have a genuine interest in AI safety and responsible deployment.

Responsibilities

• Review applications and eligibility for trusted and special access programs, verifying organizational identity, stated use cases, and account history, documenting a clear rationale for each decision.
• Provision, suspend, and revoke access according to runbooks, keeping case records complete and current the same day.
• Investigate reports of compromised accounts and leaked credentials, collaborate with affected customers on remediation, and elevate patterns to detection and engineering partners.
• Review appeals of enforcement actions, gather context from internal teams and public sources, and recommend upholding or reversing decisions with a clear rationale.
• Investigate ban evasion and clusters of related accounts, weigh signals to decide if accounts belong to the same actor and next steps.
• Draft customer notices and coordinate with account teams on approvals, violations, appeals outcomes, and access changes.
• Flag gaps between criteria and real cases, and suggest improvements to runbooks and tooling.
• Support surge volume during launches and expansions, and report on queue size, turnaround, and outcomes.

Skills

Trust & Safety

Decision making

Policy documentation

Audit trail

Communication

Job description

Safeguards Account Access & Enforcements Specialist

• Type: Contract
• Job 12492

As a Safeguards Access & Enforcement Operations Specialist, you'll help run the human review behind how Anthropic grants access to its most capable models and features, and how we respond when accounts are compromised, misused, or wrongly actioned. You'll make case-by-case decisions against detailed written policy, document your reasoning, and communicate outcomes to customers and internal partners.

The work is high-stakes and changes quickly, so it rewards precision, sound judgment, and comfort applying rules that evolve often. Important context for this role: In this position you may be exposed to sensitive content, including material related to cybersecurity misuse and other high-risk domains.

Responsibilities

• Review applications and eligibility for trusted and special access programs, including requests tied to sensitive research domains and to regions where access is restricted, verifying organizational identity, stated use cases, and account history, and documenting a clear rationale for each decision
• Provision, suspend, and revoke access according to runbooks, keeping case records complete and current the same day
• Investigate reports of compromised accounts and leaked credentials, work with affected customers on remediation, and elevate patterns to detection and engineering partners
• Review appeals of enforcement actions, gather context from internal teams and public sources, and recommend upholding or reversing decisions with a clear rationale
• Investigate ban evasion and clusters of related accounts, weighing multiple signals to decide whether accounts belong to the same actor and what action follows
• Draft customer notices and coordinate with account teams on approvals, violations, appeals outcomes, and access changes
• Flag gaps between written criteria and real cases, and suggest improvements to runbooks, review checklists, and tooling
• Support surge volume during launches and expansions, and report regularly on queue size, turnaround, and outcomes

You may be a good fit if you

• Have 2+ years in trust & safety, fraud, risk, compliance, KYC/KYB, or another operations role where you made decisions against written policy
• Can hold many detailed, evolving rules in your head and apply them consistently
• Write clear, concise decision rationales and customer-facing communications
• Keep meticulous records and understand that every decision needs an audit trail
• Know when a case is clear-cut and when it needs escalation, and elevate with a crisp summary
• Learn quickly and adapt as new processes are stood up and revised
• Work well both independently and with policy, enforcement, engineering, and sales partners
• Have a genuine interest in AI safety and responsible deployment

Strong candidates may also have

• A cybersecurity background (security operations, incident response, or threat intelligence), especially for account compromise and security-related access decisions
• Experience with business verification, due diligence, sanctions screening, or export controls
• Experience with appeals, recidivism, or linked-account investigations, including SQL or link-analysis tools
• Familiarity with AI or ML research, enough to judge whether a stated research use case is credible
• Experience running enterprise or partner access programs, including onboarding and offboarding
• Experience using AI tools to scale review workflows

Our commitment to diversity

We encourage applications from candidates of all backgrounds. Research shows that people from underrepresented groups are more likely to experience imposter syndrome.

About Anthropic

Anthropic is a public benefit corporation headquartered in San Francisco, dedicated to ensuring that artificial intelligence systems are safe and beneficial to humanity. We are a team of researchers, engineers, and policy experts working together to develop responsible AI technologies.