Plagiarism and AI at the University of Alberta: the 8 and 9 notations, and when they come off

The short answer

At the University of Alberta a grade sanction for academic misconduct may bring a transcript notation coded ‘8’ or ‘9’ — and that notation comes off by itself two years after the end of the term, with no petition required. That is the most precise published notation rule in this series, and it is worth knowing exactly, because a lot of what students fear about their transcript is vaguer than the actual rule.

The second thing to know is where AI sits in the policy. It is not filed under plagiarism. It is filed under contract cheating, alongside translation software.

A new policy, from 1 September 2024

Like York, U of A replaced its old instrument. The Student Academic Integrity Policy is the current one, and it is what governs academic misconduct now — not the Code of Student Behaviour, which older pages and older advice still point at.

The dates and the approving bodies are printed on the document:

“Original Approval Date: September 1, 2024 · Most Recent Approval: September 1, 2024 · Editorial Amendments: August 1, 2025”

“Office of Administrative Responsibility: Student Success and Experience (SSE) · Office of Accountability: Office of the Provost and Vice-President (Academic) · Approving Authority: General Faculties Council”

The amendments since have been editorial only — “Formatting and links” in March 2025, “Formatting, office name and links” in August 2025 — so the substance is the September 2024 text.

The authority behind it is statutory: section 31 of the Alberta Post-secondary Learning Act gives General Faculties Council “general supervision of student affairs including authority over student discipline”, with a power to delegate that GFC has used.

One structural point that will save you time. The policy itself is short; the offences and the sanctions live in appendices. You need four documents to see the whole picture: the Policy, Appendix A (Academic Misconduct), Appendix B (Sanction Descriptions and Impact), and the Student Academic Misconduct Procedure — with a fifth, the Appeal Procedure, if a decision goes against you. Appendices A and B were approved by a GFC committee, the Committee on the Learning Environment and Student Affairs (CLESA), and both carry 1 September 2024 as their original approval date.

The policy is reviewed “from time to time as necessary, but at a minimum every five years”.

AI as contract cheating — with translation software

Most universities reach AI through “unauthorized aid” or a plagiarism clause. U of A puts it somewhere else entirely, and the placement changes how the rule reads. Appendix A, § 3:

“3. CONTRACT CHEATING

Using a service, company, website, or application to a. complete, in whole or in part, any course element, or any other academic and/or scholarly activity, which the student is required to complete on their own; or b. commit any other violation of this policy.

This includes misuse, for academic advantage, of sites or tools, including artificial intelligence applications, translation software or sites, and tutorial services, which claim to support student learning.”

“Contract cheating” usually means paying someone to write your essay. By filing AI there, U of A puts a chatbot in the same category as an essay mill. That is a strong framing, and it is why a student searching the policy for “artificial intelligence” under plagiarism finds nothing and wrongly concludes the policy is silent.

Two things soften it, and both are in the text. The clause turns on misuse, for academic advantage — not on use. And the second limb is about work “which the student is required to complete on their own”, which is a course-level question. Neither makes the rule loose, but neither makes ordinary permitted use an offence.

The naming of translation software deserves a paragraph of its own. If you are writing in a second language, a translation tool may feel like a dictionary, not a shortcut. U of A has named it in the same sentence as AI, and that means you cannot assume it sits outside the rule. This does not mean translation is banned — the test is still misuse for academic advantage on work you were required to do alone — but it does mean it is a question to put to your instructor before you rely on it, and to get an answer to in writing. Our page on AI detectors and non-native speakers covers the related problem: second-language writing already draws suspicion it has not earned.

The other offences in Appendix A are worth a quick read, because two of them catch AI-adjacent behaviour without ever naming a tool.

Unauthorised collaboration(§ 4) includes “Submitting any assignment as the student’s own work which contains i. material generated by anyone other than the student named on the assignment, and ii. without acknowledgement and the express permission of the instructor.” Note the conjunction. Acknowledging is not enough on its own where permission was also needed — you need both.

Misrepresentation (§ 5.b) covers including “a reference to a source the student knows to contain fabricated claims […] or a fabricated reference to a source in any academic submission for assessment or credit.” That is the hallucinated-citation problem, written without naming AI, and it is the single easiest way to turn an honest use of a chatbot into a misconduct case. Check every reference a tool hands you — how to verify AI-generated citations is a ten-minute routine that removes the risk entirely.

Plagiarism itself (§ 1) is defined conventionally: “Representing the words, ideas, images, data or other analogous work of another individual or other source as the student’s own”. And unauthorised resubmission (§ 6) makes reusing your own earlier work an offence without “the written approval of the course Instructor”.

The notations 8 and 9, and the two-year clock

This is the part of U of A’s rules that no other institution in this series matches for precision. Appendix B § 3 deals with grade sanctions and what may accompany them:

“a. Marks for an assessment(s) or the grade for a course may be reduced as a sanction for Academic Misconduct. The final grade may or may not be accompanied by a transcript notation indicating Inappropriate Academic Behaviour.”

Then the rule that matters:

“d. The transcript notation indicating Inappropriate Academic Behaviour will remain on the student’s transcript for a period of 2 years from the end date of the term for the relevant course. After that time, the notation is removed. The following notations apply: i. ‘8’ for undergraduate students, or for graduate students where the resulting grade is a passing grade; ii. ‘9’ for graduate students where the resulting grade is a failing grade.”

Three things follow, and each of them is genuinely reassuring.

The notation is a numeral. It is not a sentence, not the word “misconduct”, not a description of what happened. A reader who does not know the U of A coding scheme learns very little from it.

It expires by itself. Two years from the end of the term, and then it is removed. You do not petition, you do not apply, you do not argue — the clock runs on its own. Compare UBC, where a notation is removed only on application to the President, who may decline to permit the application at all, or York, where a five-year petition is the route.

It is not automatic. A grade sanction “may or may not be accompanied” by the notation. That means whether one is imposed is something a decision-maker decides — which makes it something you can address while the decision is still open. Sanctions under this policy are explicitly meant to be proportionate: § 8.c requires them to be “proportionate based on the nature of the violation and the desire to accomplish the above aims through the least restrictive means.” That sentence is a reasonable thing to quote when a proposed sanction feels heavier than the conduct.

The rest of the scale

Policy § 8.d lists the sanctions in order: reprimand; academic integrity conditions; grade sanctions; refusal to consider current and/or future applications; rescission of an admission offer; suspension from academic program; expulsion; suspension of a degree; rescission of a degree.

Who may impose what is published, and it is useful to know where you are in the chain. § 8.e: the Faculty Decision-Maker can impose a reprimand, integrity conditions, grade sanctions, refusal of applications, and rescission of an admission offer — nothing heavier. Student Conduct Officers and the Student Misconduct Appeal Panel can impose anything on the list. So if your matter has moved to a Student Conduct Officer, the range of possible outcomes has widened, and that is the moment to get advice.

Suspension(Appendix B § 6) is “a complete withdrawal from the university, the student’s program in the university, and all university activities for a specified period of time, to a maximum of three years.” It is noted on your central academic record, your transcript and your faculty file during the suspension — and “At the discretion of the Student Conduct Officer, the suspension may be noted on the transcript for a further period of up to three years after the end of the suspension.” Under twelve months and you may re-enrol in your program; twelve months or more and you must apply for readmission. Coursework done anywhere else during a suspension will not be accepted for credit.

Expulsion (§ 7) is “a complete withdrawal from the university for an indefinite period of time”, noted on the transcript “in perpetuity”. After no less than four years a student may petition the Provost for permission to apply for admission; “The Provost’s decision is final and not subject to appeal.”

Degree suspension and rescission(§§ 8–9) are the heaviest. A suspended degree is removed from your record and shown as suspended until you meet the university’s requirements to clear it — and if you do, “the original award will be restored to the student’s central academic record with the original date. The record of the suspension will be removed from the transcript.” Rescission is permanent deletion of the award. And U of A publishes one consequence that no other institution in this series spells out: “The Student Conduct Officer may recommend to the Provost that the university publish notification of the rescission of the degree. In the case of a professional degree, this would include notification to the appropriate professional body.”

Integrity conditions: appeal at the start or not at all

One sanction on that list works differently from the others, and it contains a trap. Appendix B § 2.a:

“Academic integrity conditions is a sanction that has two components: i. the sanction requires that students satisfy specific conditions or restrictions, not to exceed the duration of the student’s program, and ii. provides for a specific secondary sanction that will be imposed on the student in the event the student does not satisfy those specific conditions or restrictions.”

So it is probation with a pre-agreed consequence attached. The trap is in § 2.f:

“Any appeal of the academic integrity conditions, including the secondary sanction, must be made at the time the initial sanction is imposed and within the time limits set out in the Student Academic Misconduct Appeal Procedure. If the conditions are not met, no further appeal is available when the secondary sanction is imposed.”

Read that twice. If you are offered integrity conditions and the secondary sanction attached to them looks disproportionate, the moment to challenge it is now — not later, when it lands. A student who accepts the conditions thinking “I will deal with that if it happens” has already given up the appeal. This is the single most expensive misreading available in the U of A policy.

The right not to be made to use AI

U of A’s Centre for Teaching and Learning publishes one sentence that creates a student right, and it is easy to misread — so we are going to be careful with it. The sentence, inside the description of the highest level of the Artificial Intelligence Assessment Scale:

“Note that the U of A requires instructors to offer a no-AI alternative to students who object to using AI.”

What that gives you is a right not to be required to use AI in your coursework. If an assignment is built around using a generative tool and you object — on ethical, environmental, religious, privacy or any other grounds — an alternative has to be available.

What it is not: it is not an opt-out from having your work checked. It says nothing about plagiarism scanning, nothing about detection tools, and nothing about where your submission is stored. We found no published opt-out or alternative-submission right of that kind anywhere at U of A. If you have read that the University of Waterloo publishes one — it does, as a compulsory clause in every course outline, described on our Waterloo page — do not carry that assumption across. Different university, different rule.

The page carrying that sentence has no publication or revision date, which we note because it is the kind of one-line right that is worth capturing a copy of if you intend to rely on it.

On the privacy side, U of A is the only institution in this series to name its provincial statute inside the integrity policy itself. Section 9.a: “Any use and disclosure of personal information contemplated in this policy and the related procedures will be in accordance with the Alberta Freedom of Information and Protection of Privacy Act.” Be clear about the scope, though: that governs the records of the misconduct process. It is not a statement about submitting student work to a third-party checking service.

Section 10 also puts confidentiality obligations on you. Parties and witnesses “should not make public another individual’s personal information that they learn solely through any university process”, which rules out posting other people’s details or circulating confidential university documents. You are not silenced about your own experience: “The university does not prohibit parties and witnesses from speaking about their own experiences”.

Acknowledgement, and the form some courses use

There is no required institution-wide disclosure form or wording at U of A that we could find. What the policy does is make acknowledgement the operative test in two places — the “misuse, for academic advantage” language in the contract-cheating clause, and the acknowledgement-and-permission conjunction in the collaboration clause.

The Centre for Teaching and Learning publishes model syllabus statements, and one of them names a specific artefact:

“All graded assignments must be accompanied by the GenAI Use: Acknowledgement and Reflection form, whether you have used AI or not.”

That is a stronger disclosure mechanic than most universities publish — but note what it is. It appears inside a sample statement offered to instructors, not a university mandate. Whether it applies to you depends on whether your instructor adopted that statement. If your syllabus mentions the form, use it every time, including on assignments where you used nothing.

The same sample statement carries a warning we would repeat verbatim:

“AI is prone to fabrication (factual inaccuracies). Review outputs carefully and validate using trusted sources. You are responsible for any errors or omissions the AI tool provides that you fail to identify and resolve.”

And the model statement for courses where AI is not allowed draws the line in a way worth knowing: “You are welcome to use AI tools as personal learning supports. However AI use is prohibited in assessed work, including pass/fail assignments, discussion boards, graded assignments and exams.” Learning support, yes; assessed work, no. If your course uses that statement, that is the boundary.

Where a tool is permitted, U of A points at the Library’s “How to Cite AI” guidance for reference formats, and warns that “Failure to acknowledge AI use may be considered cheating”. Our guide to writing an AI disclosure statement covers what a defensible acknowledgement contains.

One more U of A specific: the institutionally approved AI platform is Gemini, through Google Workspace. “Our institutional agreement with Google includes security and privacy guarantees that you may not have with a private subscription to another platform — therefore, we recommend using Gemini AI tools for your university work.” A separate set of Artificial Intelligence Data Safety Guidelines sets out which categories of institutional data — Restricted, Confidential, Protected, Unrestricted — may be shared with AI tools at all. If you are handling research data, read those before you paste anything anywhere.

What U of A does not publish about detection

We have to be straightforward here, because the honest answer is an absence.

U of A does not name a plagiarism or text-matching product anywhere we could reach. No Turnitin, no alternative, nothing. Every other institution in this series names its tool on a public page. If you want to know what checks your submission at U of A, you will have to ask your instructor or your faculty — we cannot tell you from published sources.

And no AI-detection position was found: neither on nor off, neither recommended nor discouraged. We checked the pages where the other four universities put theirs — “Using Artificial Intelligence at the U of A”, the “Framework for the Responsible Use of AI”, and the Centre for Teaching and Learning’s generative-AI resources. The CTL pages loaded and contain no reference to detection software at all.

Two caveats on that, in both directions. Parts of ualberta.ca sit behind a bot-challenge layer that blocked automated retrieval during our reading, so some documents were unreadable rather than absent — this is not proof that U of A has published nothing. The policy PDFs, by contrast, retrieved cleanly, so everything quoted from the policy and its appendices on this page is solid. Treat the detection question as open, and if you need an answer, ask your faculty directly rather than assuming.

What the CTL does publish, instead of a detection position, is an argument that detection is the wrong approach:

“Unintended consequences: While it may be simple to forbid the use of artificial intelligence, it is difficult to detect or prove whether a student has used it. Prohibiting it might simply drive the practice underground, leaving students to use these tools without your guidance.”

“Level 1 assessment activities should be proctored or used only for low-stakes formative assessment because unauthorized AI use may be undetectable.”

It is also sceptical of disclosure as a fix: “instructors should not treat disclosure statements as an assurance of academic integrity”, citing Corbin, Dawson and Liu’s 2025 argument that “disclosures alone do not secure an assessment; structural changes to assessment design are essential.”

And it names the group most likely to be hurt by a blanket ban: “Some neurodivergent students and those facing language barriers rely on support from AI tools. A strict ban may disproportionately affect these learners.” None of that is a rule you can rely on — it is guidance to instructors — but it tells you the institutional temperature, which is not one of detector-led enforcement.

Nor could we find any published student self-check route at U of A. If your instructor offers you a report on your own submission, that is a local arrangement, not a documented university service.

For what it is worth on our side: if you use our published accuracy work or anyone else’s, remember the limit U of A’s own teaching centre states — no tool proves what a person did.

How a case runs, and the price of silence

The first decision is the Dean’s, and it is not appealable. Policy § 5:

“The Dean has sole discretion to determine whether reports of academic misconduct will be addressed through non-disciplinary accountability options or through the complaint process. The Dean’s decision is final and binding.”

The non-disciplinary route is real and worth knowing about: “Voluntary, collaborative facilitated processes to explore personal accountability options outside of a complaint […] remedial, restorative, and other facilitated processes”. You cannot demand it, but it exists.

From there the chain runs: instructor report → Dean → Faculty Decision-Maker → Student Conduct Officer (for sanctions beyond the FDM’s authority) → Student Misconduct Appeal Panel.

The standard is the balance of probabilities, and unusually the policy defines it for you (§ 7.b): “This standard requires that it is more likely than not, based on the available evidence, that the student was in violation of this policy.”

The evidence rules are loose, and § 7.c says so openly: the process “is not bound by the same rules of evidence that apply in a court of law. Decision makers […] may accept and consider any credible information that is relevant to the complaint including, for example, hearsay evidence.”

Which leads to the clause every U of A student should read before deciding to stay quiet, § 7.d:

“Students alleged to have engaged in academic misconduct are encouraged to but are not obligated or required to participate in a complaint process. However, where students choose not to participate and/or decline to provide information, they may be prohibited from later attempting to provide that information or evidence to university decision-makers under the Student Academic Misconduct Procedure and the Student Academic Misconduct Appeal Procedure.”

You are not obliged to take part. But evidence you hold back now may be shut out later, including on appeal. If you have drafts, notes, version history or anything else that shows your process, the time to put it forward is the first time you are asked. Our guides on what to do when you are accused of using AI and how to prove you wrote it yourself are both written for that first response.

If a detector score is what started this, our guide to an AI detector flagging your thesis sets out what that number does and does not establish on its own.

Against that, U of A publishes the fullest list of procedural rights in this series, at § 2.g. You have the right to an impartial and unbiased decision-maker; to have your case decided within a reasonable time; to timely communication; to be accompanied by an advisor throughout any investigation, meeting or hearing, “and to be advised of these rights”; to reasonable disclosure of the allegations; to a reasonable opportunity to respond; to reasonable notice of the time, place and nature of any hearing; to “an opportunity to respond to or explain any evidence that does not support their account of events”; to provide evidence and suggest witnesses or lines of inquiry; to written reasons for any decision; and to reasonable accommodation under the Accommodation Policy.

The advisor can be a lawyer: “Assistance may be provided by the Office of the Student Ombuds, Student Legal Services, legal counsel or another advisor chosen by the student.” The Student Ombuds and Student Legal Services are free and internal, and they are the first call, not the last.

U of A also requires the people deciding your case to be trained, § 11: everyone administering, advising on or adjudicating a complaint must have training “which will include training in procedural fairness, and training to recognize and mitigate biases in processes and decision-making”. Appeal panel members are additionally trained “in the standard of review applicable on appeals”.

On published case data: the 2024 policy contains no annual-reporting clause, and we found no published counts under it. This page therefore quotes no numbers for U of A.

Before you submit

• Ask what checks your submission, because U of A does not say. No text-matching product is named on any page we could reach; your instructor or faculty is the only source.
• Treat translation tools as a question, not an assumption. Appendix A § 3 names translation software alongside AI applications.
• Check every reference a tool gives you. A fabricated citation is misrepresentation under Appendix A § 5.b, with or without AI in the picture.
• Get written permission before reusing your own earlier work. Unauthorised resubmission needs “the written approval of the course Instructor”.
• If you are offered academic integrity conditions, appeal at that moment or not at all. The secondary sanction cannot be appealed when it later lands.
• If you are asked for your account, give it. Evidence withheld now may be excluded later, including on appeal.

If you want a private read on your own writing before you hand it in, our AI check is free up to 1,500 words, with no account and no name. Our published error rates are measured on English and German corpora, and no detector — ours included — produces proof.

Sources

• University of Alberta (UAPPOL): Student Academic Integrity Policy — Approving Authority: General Faculties Council; original approval 1 September 2024; editorial amendments 1 August 2025.
• University of Alberta (UAPPOL): Appendix A: Academic Misconduct — approving authority GFC Committee on the Learning Environment and Student Affairs (CLESA); original approval 1 September 2024. The contract-cheating clause is § 3.
• University of Alberta (UAPPOL): Appendix B: Sanction Descriptions and Impact — approver CLESA; original approval 1 September 2024. The ‘8’/‘9’ notations are § 3.d.
• University of Alberta (UAPPOL): Student Academic Misconduct Procedure — read together with the Student Academic Misconduct Appeal Procedure.
• University of Alberta Centre for Teaching and Learning: AI and assessment — the no-AI alternative requirement and the undetectability argument. No publication or revision date on the page.
• University of Alberta Centre for Teaching and Learning: AI in teaching and learning — the sample syllabus statements and the GenAI Use: Acknowledgement and Reflection form. No date on the page.
• University of Alberta: Using Artificial Intelligence at the U of A — the Framework and the Data Safety Guidelines; Gemini as the approved platform. No date on the page.

This page is orientation, not legal advice. What binds you is the Student Academic Integrity Policy with Appendices A and B, the Student Academic Misconduct Procedure and its Appeal Procedure, together with your course syllabus and your instructor’s stated expectations. Where U of A publishes no rule — on which checking tool it uses, on AI detection, or on opting out of scanning — this page says so rather than filling the gap, and notes that parts of the university’s site were unreadable to us rather than empty.