Specimen — a worked example on a fictional course. Not a real certification.
Sample review report
AI-Readiness Score
78
Decision
Certified
Gates A.1 · B.5 · E.2 — all met
Certified requires 70 with all gates at anchor 2. Distinction (85, gates at anchor 3, all domains ≥ 60) was not reached: overall 78, and Domain F scored 63.
Seven of the 22 criteria are shown here to keep the specimen readable; a real report covers all of them. Every finding cites the clause it rests on — verify each against the published Standard.
The syllabus contains a titled AI-use policy covering all four assessed components, stated affirmatively (A.1.1–A.1.3 satisfied). The rationale clause recommended by A.1.4 is present for essays but absent for the problem sets.
Disclosure is required but underspecified: the policy names no format or location (A.3.1) and does not state whether honest disclosure of permitted use can affect the grade (A.3.3) — the omission most likely to teach students that honesty is risky.
Syllabus text
Any use of AI tools must be disclosed.
Suggested rewrite
If you used AI on a permitted task, add a half-page appendix titled “AI disclosure”: which system, for what purpose, and to what extent. Honest disclosure of permitted use will never reduce your grade.
The mix includes an invigilated midterm (25%) and an oral defense of the term paper (15%), satisfying B.1.1–B.1.2: a student cannot pass on unverified take-home artifacts alone. B.1.3 is partially met — the weekly memos have no verification moment.
The integrity section references detection software as primary evidence, conflicting with B.5.1. A due-process path exists (meeting, then committee referral), satisfying B.5.3. The suggested rewrite reframes detector output as indicative only.
Syllabus text
Submissions flagged by our AI-detection software will be treated as violations of the academic integrity policy.
Suggested rewrite
Detection tools may prompt a conversation, but their output alone is never treated as proof. If concerns arise, you will be invited to discuss the work and show your process materials (drafts, history, notes) before any referral under the university's integrity procedure.
Exemplary. The syllabus prohibits entering interview transcripts into external AI systems, names the institutionally provisioned alternative, and lists what may never be shared (E.2.1–E.2.3 all satisfied). This treatment could serve as a model.
The syllabus requires student disclosure but says nothing about the course's own AI use (F.2.1). Feedback on weekly memos is AI-assisted per the instructor's intake notes, which makes the silence a material asymmetry (F.2.2).
Syllabus text
— (no statement present)
Suggested rewrite
AI use by the course: weekly memo feedback is drafted with AI assistance and reviewed by the instructor before release; grades are assigned by the instructor alone. If you believe feedback misread your work, raise it in office hours for human re-review.
Advisory (unscored in v2026.2): the course includes one week on computational text analysis but does not yet engage how AI is reshaping comparative research practice (D.1.1 partially met). Reported for the professor's consideration; no effect on the score.
In the live product, every suggestion above arrives as an accept/reject decision for the professor, with follow-up questions answered in place. Nothing is applied without explicit acceptance, and the final decision is computed mechanically from the published conformance model.