What is a stakes profile?
A candidate's report under AIRS-J/1 lists the stakes profiles assessed. These seven entries give the Standard's own wording for that line and the words around it, each with its section, beside a plain reading that adds no claim to it.
Judgment is how well a person decides once AI has given the answer. AIRS-J/1 is the standard AIR APAC publishes for judgment in AI-enabled work.
A candidate's report under AIRS-J/1 lists "the stakes profiles assessed" (Section 12.5). The seven entries below explain that line and the words that go with it. Each entry gives the Standard's own wording first, with its section. A plain reading follows. The plain reading adds no claim to the Standard. Where the two differ, the Standard governs.
1. Stakes profile
The Standard's wording. "The consequence families in play in a given assessment instance (Section 6)." (Section 3)
Plain reading. A stakes profile names the kinds of harm that one assessment instance puts at stake (Section 6.1). The specification of the official instrument sets out each scenario's stakes profile (Annex F).
2. Primary and secondary consequence family
The Standard's wording. Every assessment instance declares "one primary consequence family and at most one secondary, drawn from the canonical set below" (Section 6.1). Section 3 has no row for these terms.
Plain reading. Each scenario names one main kind of harm, the primary family. It may name one more, the secondary family, and never more than that. Both come from the fixed list of six in Section 6.1.
The six families, by code: C1 People safety and welfare. C2 Financial and fiduciary loss. C3 Operational and service continuity. C4 Information and privacy harm, including wrong or misused records about identifiable people. C5 Trust, reputation and legitimacy (C5a commercial; C5b institutional and public). C6 Environmental and community-system harm.
3. When a family belongs in a profile
The Standard's wording. Section 6.3 sets the eligibility rule. Section 3 has no row for it. "A consequence family belongs in a stakes profile only where the instance":
- "identifies an affected bearer";
- "provides a credible pathway from the candidate's decision to the harm";
- "makes severity or reversibility relevant";
- "requires the candidate to accept, reduce, transfer or monitor that harm".
"Mentioning a policy, a dataset or a stakeholder is not sufficient." (6.3)
Plain reading. A scenario cannot claim a kind of harm that it only mentions. The scenario must name someone who would be affected by the harm. The candidate's decision must be able to lead to that harm. How serious the harm is, or whether it can be undone, must matter. The person must have to deal with the harm: accept it, reduce it, pass it to another party, or keep watch on it.
4. Contextual qualifier
The Standard's wording. "Duty of care and vulnerability, public accountability, dependency and cascade, and binding legal or contractual constraint may be recorded as qualifiers on an instance. They are not scored and are not consequence families." (Section 6.5) Section 3 has no row for this term.
Plain reading. A qualifier is a note on the situation in a scenario. There are four:
- someone in the situation is owed a duty of care, or is vulnerable;
- the decision is answerable to the public;
- others depend on the outcome, so a fault can spread;
- a law or a contract binds the decision.
A qualifier is not one of the six families. It is not scored.
5. Like stakes
The Standard's wording. "Judgment scores are not comparable across situations of different consequence." (Section 6) "Consequence families are not scored. They are recorded so that observations are compared within like stakes." (6.2) Section 3 has no row for this term.
Plain reading. The families are recorded so that an observation is compared with observations made under like stakes (6.2). A careful check in a low-stakes exercise and a careful check where a person could be harmed are not the same behaviour (Section 6). The families in it are not scored (6.2). A stakes profile records the kinds of harm at stake and gives no score to the person or the scenario.
6. "The stakes profiles assessed" on a report
The Standard's wording. Section 3 has no row for this phrase. The Standard says:
- Every candidate receives "the stakes profiles assessed" with the rest of the report. "This is provided whether or not the candidate certifies." (12.5)
- "A candidate's report states the stakes profiles under which their observations were generated." (6.6)
- "An employer reading a Certified result can see the consequence range it was earned across." (6.6)
- The register "does not record employer, dimension scores, stakes profiles, or any transcript content." (14.1)
Plain reading. The line lists the kinds of harm that were at stake in the scenarios where the person's observations were made. An employer that reads a Certified result sees from it the range of consequence the result was earned across (6.6). The line gives no score for any family. It does not say how hard the assessment was. The public register does not record it (14.1). A check of a credential code in the register therefore does not show it.
7. A localised scenario keeps its stakes profile
The Standard's wording. Section 3 has no row for this rule. The Standard says:
- "Scenario setting, sector, role titles, organisational norms, regulatory context and language may be localised. The decision points, the dimensions they elicit, the stakes profile and the anchors applied may not." (8.5)
- For a candidate's access support: "Approved supports must preserve the construct, decision points, stakes profile and assessment integrity" (8.9).
Plain reading. A scenario may be adapted to another setting, sector or language. The kinds of harm it puts at stake stay the same. A localised scenario needs AIR APAC approval before it is used in certification (8.5). An approved access support for a candidate also keeps the stakes profile the same.
Related rule
Whether a set of stakes profiles is wide enough for certification is a separate rule, stakes coverage (Section 6.4). It is rule 5 of the six rules in Section 12.2, where "Failure of any one results in Not Certified."
What these entries do not claim
They explain words of AIRS-J/1 v2.3 and do not amend its text, hold no assessment scenario, and say nothing about any person's result.
AIR APAC publishes the standard for judgment in AI-enabled work, certifies people against it through its official instrument, and stands behind every result.