What You Need to Know
Accountability means answering for work when it goes wrong. Disclosure means telling people AI helped make it. A stakeholder is anyone the work affects. Whether an output is correct is Domain 2's question. This objective asks who owns the output, who deserved to know how it was made, and who gets described by it. Unlike most objectives in this domain, it has little documentation.
Take a family-run takeaway using Claude for its newsletter, its website order chat and the weekly specials the head cook used to write.
Accountability
The person whose name is on the work owns it: the email's sender, the report's publisher, the letter's signer. Saying Claude drafted it describes the production method, like saying an assistant typed it, and moves no responsibility. The Usage Policy writes this into its high-risk requirements. When Claude's output becomes advice or a decision in one of its High-risk Areas, such as legal or medical advice, a qualified person must "remain accountable for what is delivered or decided". The AI Fluency framework calls the same duty Diligence.
If the newsletter prints a wrong price under the owner's name, the owner owes the correction. Blaming Claude, or whoever configured the Project, only explains how the error happened. Under pressure that is tempting, so drill the rule until it is reflexive.
When Disclosure Is Owed
The Usage Policy sets one categorical rule. Consumer-facing chatbots and external-facing AI agents must tell users they are interacting with AI rather than a human. This must happen at a minimum at the beginning of each chat session or in the product interface. The takeaway's order chat can open each conversation by saying it is AI, or carry a label in the chat window. A line in a privacy notice elsewhere on the site is neither.
Everything else runs on proportion, and both extremes are wrong. Nobody declares AI assistance on a routine internal email like the staff rota, and declaring everything is over-caution. Disclosure is owed when:
- the reader's trust depends on who did the thinking or the judging: assessed work, a professional opinion
- a claim of authorship would be false
- a policy or client agreement requires it
The test is what the reader relies on. If it is the information, disclosure adds little. If it is the fact that you did the thinking, it is owed. A personal message such as condolence, or thanks meant as your own feeling, is different. Its reader relies on you having done the feeling, and a disclosure note cannot give that back, so you write it yourself. An organisation's routine thank-you letters are ordinary drafting work. The newsletter's prices need no note. The owner writes the thank-you card for a retiring driver without Claude.
Fairness Is a Different Review from Accuracy
An output can pass every accuracy check and still be unfair. Descriptions of groups, communities, applicants or service users can repeat assumptions the facts never required. Say the newsletter's delivery section frames one estate as a problem, a village as declining and older customers as resistant to change. Every delivery time is correct, yet those described still see the unfairness at once. Checking numbers misses it, because it sits in the framing. It needs a different reviewer: someone who knows those customers, reading for how they are described. The Usage Policy also prohibits using Claude to critique anyone's body shape or size, as promoting unhealthy body image.
Domain 2 owns spotting bias as an evaluation skill. This objective owns the responsibility half: whose job it is to catch unfairness in work about people before it publishes, and what the publisher owes when it gets through.
Colleagues Are Stakeholders Too
The stakeholder most candidates never consider is the person whose work changes. When Claude takes over part of a process, the colleagues doing that work need to know what changed and why. Finding out a shared task was automated without a word damages the trust disclosure protects. So the owner tells the head cook before Claude's first specials board goes up. Using Claude to produce a colleague's contribution without telling them is an ethical question of honesty between colleagues, not output quality. Examples: appraisal comments they were meant to write, a section they were meant to own.
In the exam
Stems ask who owns an output, who should have known, who is described, or whose work changed. The keyed answer keeps accountability with the name on the work. Distractors spread it onto the tool or the Project's configurer, or offer a blanket declare-everything rule, which is over-caution.
The exam tell
An option whose substance is checking, cross-referencing or validating answers a Domain 2 objective. A 6.4 key is always about ownership, disclosure, fairness or colleagues.
Where These Rules Come From
Extending the high-risk sentence to everyday workplace outputs is the exam guide's objective read through ordinary professional practice. The same two grounds carry the disclosure framework beyond the chatbot rule, and the section on colleagues; no first-party document covers either.
Exam Traps
Transferring accountability to the tool or its configurer
Answering a challenge with the explanation that Claude produced the figure, or that someone else set up the Project, describes how the error was made and leaves untouched who owes the correction. Responsibility attaches to the name on the work, and blame-the-tool options are never keyed.
Using disclosure as a shield or as a blanket rule
An AI-assistance note added to soften ownership of an error misuses disclosure, and declaring assistance on every routine output is over-caution that drowns the cases where it genuinely matters. Disclosure is about what the audience is relying on, not a ritual or a deflection.
Signing off work about people once the figures check out
Fairness failures survive accuracy checks by definition. The characterisation review is a separate pass by someone who knows the people described, and an option that stops at verified figures has answered the accuracy question while the ethical one is still open.
Practice Scenario
The clerk of an almshouse charity publishes the annual report to residents and donors, largely drafted by Claude from the charity's own maintenance and occupancy records. A resident spots that the report overstates how quickly repairs were completed last year, and asks the clerk who wrote it. What is the professional response?
Build Exercise
Audit One Published Piece
What you'll learn
- How to write the ownership line for a piece of AI-assisted work
- How to decide whether disclosure was owed and on what ground
- How to run a fairness pass separately from an accuracy pass
- How to identify the colleagues a piece of work affected
- Pick one piece of AI-assisted work you have already sent or published, ideally with your name on it and an audience beyond your own team
Why: The audit only teaches on work with real ownership and real readers; drafts and private notes have no stakes to examine.
You should see: One named piece: what it was, who received it, whose name was on it.
- Write its ownership line: whose name is on it, and exactly what you would say if a reader challenged a claim in it tomorrow
Why: The anchor only holds if you can say the words; drafting them now is what prevents the accountability transfer under pressure.
You should see: Two sentences: the owner, and the response to a challenge that accepts ownership before explaining anything.
- Decide whether disclosure was owed: what was the reader relying on, the information itself or the fact that you did the thinking, and did any policy or client agreement require a declaration
Why: The proportion test needs practice on real cases, because both extremes feel safe and neither is.
You should see: A yes or no with the ground named: reliance, authorship, or a policy requirement.
- Re-read every passage that describes a group of people and mark anything accurate but unfair; name who knows those people well enough to review it, and whether they did
Why: The fairness pass is a separate skill from fact-checking, and it only becomes visible when run separately.
You should see: Marked passages, or a note that none describe people; plus the name of the right reviewer.
- Note whose work the piece changed or replaced, and whether they knew; if they did not, that conversation is the output of the exercise
Why: The colleague stakeholder is the one this objective says candidates always miss, and the audit is incomplete without them.
You should see: Either a note that colleagues knew, or the name of the person who deserves the conversation and a line on what you will say.
Sources
- Usage Policy — Anthropic
- AI Fluency: Framework and Foundations — Anthropic
- Claude Certified Associate – Foundations Exam Guide — Anthropic. Beyond the chatbot disclosure rule and the high-risk responsibility sentence, no Anthropic document covers this objective. The sections on fairness and on colleagues draw on the exam guide's objective and standard professional practice.