Governance, Risk, and Responsible Use
Free Claude Certified Associate - Foundations practice — 6 questions on Governance, Risk, and Responsible Use, with explanations. No sign-up. Full 12-question mixed test →
A marketing analyst at a healthcare company wants to quickly summarize customer feedback that includes patient names, diagnoses, and insurance IDs. She plans to paste the raw feedback into her personal, free Claude.ai consumer account because it's the fastest option available. Per responsible AI and data-privacy practices, what is the BEST course of action?
Enterprise/Team agreements include contractual data protections (e.g., no training on prompts, DPA, retention controls) appropriate for regulated PHI, and de-identifying sensitive data further reduces risk regardless of tool.
A clinic is piloting Claude to help a nurse practitioner draft patient discharge instructions. Which practice aligns with Anthropic's Usage Policy for high-stakes domains like healthcare?
Anthropic's Usage Policy requires human oversight for high-stakes outputs such as medical content — Claude can assist, but a qualified professional must validate outputs before they reach a patient.
What is the primary purpose of Anthropic's Responsible Scaling Policy (RSP) and its associated AI Safety Levels (ASL)?
The RSP defines capability thresholds (increasing AI Safety Levels) with corresponding security and deployment safeguards designed to manage catastrophic risk as model capabilities scale.
A compliance officer asks whether prompts and outputs submitted through the Claude API will be used to train future Anthropic models. Which statement accurately reflects Anthropic's default commercial data policy?
Anthropic's commercial terms state that API inputs and outputs are not used for model training by default, and are retained only for a limited window (e.g., trust & safety review) unless the customer opts in or law requires otherwise.
An HR team configures Claude to automatically reject job applicants who score below a threshold on an AI-generated 'culture fit' summary, with no human reviewing rejected applications. Which of the following represents the WORST practice among these options, per responsible-AI governance?
Fully automating consequential, adverse employment decisions without human review removes accountability and can embed or amplify bias undetected — a core governance concern for high-stakes, consequential AI decisions.
A company publishes blog articles that are substantially written by Claude with minimal human editing, and does not disclose AI involvement to readers. From a responsible-AI governance standpoint, what is the main concern with this practice?
Transparency and disclosure of AI-generated content is a recognized responsible-AI governance practice, and increasingly a legal/policy requirement, to maintain reader trust and informed consent.
Ready for the real thing?
The full course has two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed answer explanations.
Start my full course on Udemy →