Quick Reference: Domain 6 — Governance, Risk, and Responsible Use

Fit is not permission. Domain 4 asks whether a task suits Claude (shape, volume, checkability); this objective asks whether it should be handed over at all. A task can be structured, voluminous and checkable, and still belong to nobody but a qualified human.

Four categories make a use inappropriate:

The substitution test: would you accept the work from a capable colleague with no professional qualification in the area, no access to the people's history, and no accountability for the outcome? Three grounds: qualification, history, accountability.

The counterweight: most work is appropriate, and the exam punishes over-caution too. When a proposal mixes appropriate and inappropriate work, the keyed answer keeps the appropriate half. Tells: the internal-use excuse is never keyed, and adding human review to an inappropriate delegation repairs nothing.

Recognition has two halves. The obvious inventory: names, addresses, dates of birth, account numbers, health information, anything about children. The under-recognised half: material sensitive because of who owns it or what was promised about it, a duty of confidence, commercially confidential figures, third-party material under contract. A pricing schedule with no personal data in it can still be the most restricted document on the desk.

The ladder, in order; the keyed answer is the lowest rung that works:

Abandonment sits below the ladder and is almost never keyed; the guide's own sample rationale says anonymisation enables the analysis.

The placebo family fails as one. Instructing Claude not to retain the data, declaring the upload confidential, deleting the conversation afterwards: all operate after the disclosure, and the policy control is about what leaves the organisation's hands, which happened at upload. The safeguard happens before the upload or not at all.

Which product are you on? Consumer and commercial products answer the training question differently, and the details change. Items key on the professional behaviour, checking the current documentation for your product and your agreement, never on a memorised retention period or default.

The workspace rule is substance, not etiquette. Three things attach to an organisation-provided workspace and reach no personal account: the organisation's agreement (the organisation decides what happens to its data; Anthropic handles it on the organisation's instructions, and commercial products do not train on inputs or outputs by default), administration (workspace settings sit with the organisation's administrators), and data handling (on Enterprise, the organisation sets custom retention timelines for its data). A paid personal plan with carefully chosen settings is still a personal account: what matters is whose agreement governs the data.

The shadow-AI response has two halves, and distractors reliably offer only the first. Do not copy the workaround, however pressing the deadline; then raise it with whoever owns the policy or the data, because the disclosure that already happened needs to be known and an unreported workaround continues. Having the colleague delete the chats is after-the-fact housekeeping, not remediation.

Policy silence is not permission. A policy that does not mention something has not answered the question; the keyed move is asking the policy owner. When policy conflicts with convenience, the policy wins; when the policy is genuinely wrong or too slow, the route is getting it changed by its owner, not working around it.

The examinable habits: know which account and plan you actually work in, know who administers it, follow the upload rules, know how to report a problem.

The accountability anchor: the person whose name is on the work owns it completely. Saying Claude drafted it is a statement about production method, like saying an assistant typed it, and transfers nothing, not to the tool, not to whoever configured the shared Project. The Usage Policy's high-risk requirements say the same: for advice or decisions in its High-risk Areas, a qualified person remains responsible for the accuracy and appropriateness of what is delivered or decided.

Disclosure runs on proportion, with one categorical case. Consumer-facing chatbots and external-facing AI agents must disclose that users are interacting with AI. Everything else follows the reliance test: what is the reader relying on? If it is the information alone, disclosure adds little; if it is the fact that you did the thinking (assessed work, a professional opinion), disclosure is owed. A condolence letter is relied on as your own feeling, which a disclosure note cannot restore, so you write it yourself. Both extremes are distractors: nobody declares AI assistance on a routine internal email, and a blanket declare-everything rule drowns the cases that matter.

Fairness is a separate review from accuracy. An output can pass every fact-check and still characterise a group unfairly; the review needs someone who knows the people, reading specifically for how they are described.

Colleagues are stakeholders. When Claude takes over part of a process, the people doing that work deserve to know what changed and why; quiet automation of a shared task is an honesty question, not a quality one.

Boundary tell: if an option's substance is checking, cross-referencing or validating, it answers Domain 2's question. A 6.4 key is about ownership, disclosure, fairness or colleagues.

Current state, checked 30 September 2026. These are the volatile facts behind this domain. They change, and none of them is worth memorising as an exam answer; what the exam rewards is the checking behaviour. Which product am I on, what does its documentation say today, and what does my organisation's agreement add?

• The Usage Policy (formerly the Acceptable Use Policy) is updated as Anthropic's technology and the associated risks evolve. By 11 October 2026 it had been reorganised around High-risk AI Recommendations in named High-risk Areas. It carries the chatbot disclosure requirement and the high-risk accountability line this domain leans on.
• Training defaults differ by product. Two live articles share the title "Is my data used for model training?": the consumer one says consumer products train subject to the user's own model-improvement setting; the commercial one says commercial products do not train by default.
• Retention periods (from "How long do you store my data?"): deleted conversations leave back-end systems within 30 days; feedback submissions are kept for 5 years; opted-in training data is retained de-identified for up to 5 years; inputs and outputs flagged for Usage Policy violations for up to 2 years, with trust and safety classification scores for up to 7 years.
• The privacy policy is effective 10 September 2026, with Anthropic Ireland Limited as the controller for the EEA, UK and Switzerland (checked 30 September 2026).