The opportunity
Anthropic's Safeguards team is responsible for enforcing our policies, protecting users, and ensuring our platform is not misused. A large and growing share of Claude usage reaches customers through our cloud partners: Amazon Bedrock, Google Cloud's Agent Platform, and Microsoft Foundry.
What you'll do
Own the end to end enforcement lifecycle on partner surfaces, from detection: and review through action, customer communication, and appeal
Define enforcement paths for each partner's operating model, including how: actions are requested, executed, and confirmed when the partner controls the account
Manage SLAs for review, escalation, and action, and evaluate performance against them on an ongoing basis
Develop and maintain SOPs and escalation guides for all enforcement: workflows, ensuring consistency across clouds
Establish red-teaming, Quality processes, and operationalize shared metrics
Stand up paired investigations with partner threat intelligence teams,: including case referrals, indicator sharing, and paging for high severity incidents
What they're looking for
- Direct trust & safety, abuse, or security operations experience at AWS, GCP,: or Azure, or at a company that ships through their marketplaces
- Experience designing incident response or escalation processes that span more than one company
- Familiarity with data access controls, audit requirements, or privacy constrained review environments
- Experience with law enforcement referrals, regulatory reporting, or compliance programs
- Background in AI safety, content moderation, fraud, or abuse detection systems
- Comfort working with data and metrics to inform operational decisions and surface trends to leadership
