Is AI Allowed in My Assignment? Check the Syllabus First

The short answer. It depends on the course, and that answer is starting to be written down where you can actually find it. From 1 September 2026, the University of British Columbia requires every course syllabus to include a statement on how generative AI is or is not permitted in student-submitted work, under its Senate Policy V-130. That matters well beyond one Canadian university, because "was this use authorised?" is the first question in any AI misconduct case, and until now most students have had to guess the answer. If your syllabus says nothing, do not assume either way: ask the instructor in writing before you submit.

Most arguments about students and AI jump straight to detection: was this essay written by ChatGPT, and can anyone prove it? There is an earlier question that decides most real cases, and it gets far less attention. What use was actually allowed in the first place?

What UBC has done

UBC's teaching centre states the requirement plainly: effective 1 September 2026, all UBC course syllabi must include "a statement on generative artificial intelligence (GenAI) usage in student-submitted work that clarifies how the use of GenAI is permitted or not permitted". It arrives alongside a required academic integrity statement; UBC's syllabus guidance lists both among the policy statements that Senate Policy V-130, the policy governing what a syllabus at UBC must contain, makes mandatory from the same date.

Two design choices in the policy are worth noticing.

First, UBC did not write a university-wide AI rule. Its guidance says the use of GenAI tools "is a course or program-level decision", that appropriateness "differs according to disciplinary context and course learning goals", and that statements "may need to be granular to specific assessments". A literature seminar and a programming course can legitimately set opposite rules, and the same course can allow AI in one assignment and ban it in the next. The policy mandates clarity, not a position.

Second, the requirement puts the burden of clarity on the institution, not the student. The background to the policy change is telling. The Senate committee minutes behind the revision record that conversations with student senators and UBC's teaching centre "indicated that it is not standard practice for instructors to clarify their stance on AI use in their courses", and that "because permitted uses of GenAI may vary by class, a blanket statement was not feasible". Students were being held to expectations that had never been written down.

Why explicit permission is the first evidentiary step

When an AI-use allegation lands, the case runs through a sequence. Was AI use in this assessment prohibited, restricted, or allowed? If restricted, did the student's use cross the stated line? Only then does anyone reach the question of what the evidence shows.

A missing or vague policy breaks that sequence at step one. A student who used AI to structure an outline in a course that never stated a rule is in a genuinely different position from a student who ignored a written ban, and treating those two cases the same is hard to defend on appeal. It is the same principle we describe in what JCQ says about AI detection in coursework: process and stated expectations come first, and detector output is a screening signal for a human judgement, never proof on its own. Our own published testing shows why no one should want a score to carry a case alone, including how often detectors flag real student writing.

An explicit syllabus statement converts an unwinnable argument about intent into a checkable question about compliance. That protects students from retrospective rule-making, and it protects instructors too: a misconduct finding is far easier to defend when the rule the student broke was in writing before the work was set.

The UK version of the same question

For UK readers the principle is identical, but the document goes by more names. At university, the AI rule for a piece of work may sit in the module handbook or the assessment brief rather than in anything called a syllabus, so check all three before assuming there is no rule. At school level the expectation is already set nationally: JCQ puts the obligation on the centre to have an AI policy and on the student to acknowledge AI use, which we cover in what JCQ says about AI detection in coursework. Wherever the rule lives, the sequence below is the same.

If you are a student and the syllabus is silent

The uncomfortable reality is that most institutions are not where UBC is. Plenty of courses this term still have no written AI statement. If that is your situation:

• Do not read silence as permission. Academic misconduct policies predate ChatGPT, and many institutions treat unauthorised assistance as covering AI even where AI is never named.

• Ask in writing, before submitting. One email: "Does this assignment permit the use of generative AI tools, and if so for what?" The reply is your evidence if the question ever comes up later.

• Keep your process evidence. Drafts, notes, version history. In a dispute, being able to show how the work developed matters more than any detector score, a point the assessment-design case we covered makes at course scale.

If you teach

The UBC model is a reasonable template even where nothing mandates it: state a rule per course, vary it per assessment where needed, and say why. Our school AI use policy template covers the school-level version of the same exercise, and the teachers' guide to detectors covers what a detector result can and cannot tell you once a rule exists.

One caution on scope: V-130 is UBC's policy. Other universities are converging on the same need for clarity, but they still differ materially on what they permit, what disclosure they require, and what the consequences are. Check your own institution's current policy rather than assuming any of this transfers.

Sources

• UBC Centre for Teaching, Learning and Technology, Developing your generative AI syllabus statement: states the 1 September 2026 requirement under Senate Policy V-130 and the per-course, per-assessment variation guidance.

• UBC CTLT, Your course syllabus: lists the policy statements V-130 requires from 1 September 2026, including the academic integrity statement and the GenAI usage statement.

• UBC Senate Academic Policy Committee, minutes of 19 January 2026: records the V-130 revision motion, the finding that stating a stance on AI use was not standard practice, and why a blanket statement was not feasible.