Governance, Safety & Risk Management

Manage prompt injection, data handling, compliance, human oversight and responsible deployment.

1. 5.1Threat Modelling Agentic SystemsHow to threat model a Claude system that reads untrusted content and acts through tools, the attack vectors specific to agents, and how to size and shrink the blast radius before an attacker finds it.10 min
2. 5.2Prompt Injection DefencesA defence-in-depth model for prompt injection in Claude systems, from architectural separation and provenance tagging to detection, output validation and human checkpoints, with the honest limits of each layer.10 min
3. 5.3Data Privacy, Retention and ComplianceWhat an architect must know about how data flows to and is retained by the Claude platform, the controls available, and how to translate privacy and regulatory obligations into design decisions.10 min
4. 5.4Human-in-the-Loop and Oversight DesignHow to decide where humans belong in a Claude system, design approval and review mechanisms that people actually use well, and avoid the failure modes of rubber-stamping and oversight that nobody can staff.9 min
5. 5.5Responsible Deployment and Usage PolicyHow Anthropic's Usage Policy and safety behaviours shape what you may build and how, and the responsible deployment practices an architect is expected to put in place before and after launch.9 min