Organizational Governance and Escalation Paths
CoreFollow organizational AI policies and governance · Difficulty 1/5
Explanation
The AUP sets Anthropic's outer boundary, but each organization layers its own governance on top: approved tools and plans, allowed data types, required review steps, and escalation paths for exceptions. Responsible use means knowing and following those standards, not improvising a personal interpretation of what seems reasonable.
When Policy Is Unclear: Escalate, Don't Guess
When policy is unclear or a case is genuinely novel, the correct move is to escalate to the right owner rather than guessing. This is a direct parallel to the 'find the compliant path' pattern in Domain 6a: rather than picking an extreme (assume it's allowed and proceed, or ignore the policy altogether, or perversely do the opposite of what seems intended), the correct behavior routes the decision to whoever actually owns that policy.
Worked Example
An organizational AI policy is unclear about a novel use case. The best move is to escalate to the policy owner for guidance before proceeding -- not to assume it's allowed, not to ignore the policy, and not to do the opposite of what seems intended.
Common exam traps
- Improvising when a policy is unclear instead of escalating to the policy owner. 'It seemed reasonable' is not a governance process.
- Treating ambiguity as license to proceed -- an unclear policy is a signal to ask, not a green light.
Key Takeaways
- Organizations layer governance (approved tools, allowed data, review steps, escalation paths) on top of the AUP
- Responsible use means following known standards, not improvising
- When policy is unclear or the case is novel, escalate to the policy owner rather than guessing
- Exam trap: assuming ambiguity means permission, or improvising instead of escalating
Glossary Terms
A reliability pattern where the agent recognizes conditions it cannot handle autonomously and escalates to a human or higher-capability system. Escalation triggers include: conflicting data sources, low confidence scores, ambiguous requirements, or irreversible high-stakes actions.
Anthropic's outer boundary of acceptable use for Claude, with heightened requirements for high-risk and agentic scenarios. Organizations layer their own AI policy on top of it (approved tools, allowed data types, review steps, escalation paths); both must be respected simultaneously, and neither pricing pages nor the model's own self-assessment substitute for it as the authoritative standard.
Related Concepts
Prompt Injection and Treating External Content as Untrusted
Prompt injection = malicious instructions hidden inside external content (documents, web pages)
Skill Trust Evaluation: Source, Reach, Appropriateness
A Skill runs as software inside your session: it inherits whatever the session already has access to, and can act via code execution
Choosing the Compliant Path Over Abandonment or Rule-Breaking
The correct move for a borderline case is usually to find the compliant adjustment, not to proceed as-is or abandon the task