Source: UC Berkeley (2026)
KEY TAKEAWAYS
- → sixty (60) to ninety (90) days UC Berkeley's own Code of Student Conduct states that investigations 'shall be concluded promptly, generally within sixty (60) to ninety (90) days' – the longest formally published institutional window in this audit. (UC Berkeley, 2026)
- → two days The tightest student-side deadline found: NC State gives a student two days to choose how to resolve the case. (North Carolina State University, 2026)
- → 90 calendar days Brunel is the only procedure in the audit that caps the entire process end to end, at 90 calendar days from the start of the investigation. (Brunel University London, 2025)
- → 73% 73% of the 2,400-plus student misconduct cases LSU handled were AI-related, and the provost confirmed a backlog that left students waiting months. (WAFB 9 News, 2026)
- → None stated The UK ombudsman that hears these complaints publishes no data on how long providers take, so no regulator is measuring the wait either. (Office of the Independent Adjudicator, 2025)
- → five (5) days Maryland is one of the few procedures where the clock runs both ways: five days for the student to appeal, five days for the office to answer. (University of Maryland, 2026)
- → 20 working days UCL must convene a panel within 20 working days of the Stage 1 outcome and notify results within five – the tightest institution-side schedule in the audit. (University College London, 2026)
Every student flagged for AI use asks the same question, and it is never the one about detectors. It is how long this is going to last. How long does an academic misconduct case take, start to finish? The answer they get is a procedure document, and the answer they want is a date. Those are not the same thing, and the gap between them is where the anxiety lives.
So we read the documents. This article audits nineteen published academic misconduct procedures across the United States, the United Kingdom, Australia and Canada, and counts something nobody appears to have counted before: not how long a case takes, which no institution publishes, but whose deadlines are actually written down. The result is lopsided. Every procedure in the audit sets at least one firm deadline the student must meet, some as short as two days. Six of them set no deadline whatsoever for the institution to reach a decision.
That asymmetry reframes the whole experience. The wait that students describe as the system breaking down is, in most of these documents, the system running exactly as drafted. If you have just been flagged and you are still in the first day of this, what you do in the next twenty-four hours matters more than anything in this article. If you are already weeks in and wondering whether the silence is normal, read on.
1 Who Is Actually on a Clock? The American Procedures
Across ten United States procedures, every single one sets a deadline the student must meet. Only six state a deadline for the institution to reach or communicate a decision. Four bind the student and leave their own side of the process open-ended.
| INSTITUTION | INSTITUTION'S OWN DECISION DEADLINE | SOURCE |
|---|---|---|
| Indiana University Bloomington | seven (7) business days | Indiana University Bloomington |
| North Carolina State University | None stated | North Carolina State University |
| Missouri State University | None stated | Missouri State University |
| University of Wisconsin-Madison | 14 days | University of Wisconsin-Madison |
| Arizona State University (Engineering) | None stated | Arizona State University (Engineering) |
| University of Maryland | five (5) days | University of Maryland |
| CT State Community College | 15 business days | CT State Community College |
| Cornell University | 7 days | Cornell University |
| University of Memphis | None stated | University of Memphis |
| UC Berkeley | sixty (60) to ninety (90) days | UC Berkeley |
Read enough of these side by side and a drafting habit becomes obvious. These documents are written as compliance instruments, and compliance instruments point outward. The passages that carry hard numbers are the passages that create an obligation for the person being investigated: respond by this date, appeal by that date, complete this module within that window. The passages describing what the institution will do drift into a different register entirely. Hearings are scheduled as soon as practical. Cases are resolved when possible. Outcomes are communicated in a timely manner.
The exceptions are instructive precisely because they are exceptions. Maryland runs a symmetric clock, giving the student five days to appeal and the office the same five days to answer. Wisconsin-Madison commits to written results within a fixed window after the hearing. Berkeley, unusually for an American procedure, puts a number on the investigation itself. What separates these from the rest is not that their cases are simpler. It is that somebody drafting the policy decided the institution should be answerable to a date too.
None of this makes a long case improper. It does mean that if you are waiting, and your institution sits in the group that binds only you, there is probably no internal deadline being missed on your behalf. That is worth knowing before you write an email accusing anyone of delay. It is also worth knowing before you assume that silence signals something about the strength of the case against you, because what actually lands on your record is decided at the end of the process, not during the wait.
2 The Same Audit Outside the United States
The pattern reverses abroad. Of the six United Kingdom procedures audited, five commit the institution to a decision window. Both Australian procedures do. The single Canadian procedure quantifies neither side, leaving the student and the university equally unbound.
| INSTITUTION | INSTITUTION'S OWN DECISION DEADLINE | SOURCE |
|---|---|---|
| University College London (UK) | 20 working days | University College London |
| Brunel University London (UK) | 90 calendar days | Brunel University London |
| Solent University (UK) | None stated | Solent University |
| Canterbury Christ Church University (UK) | 35 working days | Canterbury Christ Church University |
| University of East London (UK) | ten working days | University of East London |
| University of West London (UK) | 20 working days | University of West London |
| Adelaide University (AU) | five working days | Adelaide University |
| James Cook University (AU) | 10 University Working Days | James Cook University |
| University of Manitoba (CA) | None stated | University of Manitoba |
British procedures read differently because they are written against a regulatory backdrop that expects them to be. They tend to specify a window for the investigation, a window for convening a panel, and a separate window for notifying the outcome, and they repeat those windows at every stage including appeal. The Australian documents follow the same habit. The practical effect is that a student at one of these institutions can open the procedure, find the stage they are in, and derive a date that ought to have passed.
The qualifiers still matter. Almost every institution-side window in this group is softened by the word normally, and none of the documents audited states what happens when the window is missed. A deadline with no consequence attached is a service standard, not a guarantee. But a service standard you can quote back is still worth considerably more than a promise to act as soon as practical, and students in these systems have something concrete to point at when a case goes quiet.
Manitoba is the honest outlier. Its guidance tells the decision-maker to give the student a deadline and to offer a timeline for the decision, without fixing either. Nothing in it is unfair on its face. It simply declines to commit anyone to anything, which leaves the student with no benchmark at all and no way to tell an ordinary delay from an abandoned file.
3 How Fast You Have to Move, and How Little Time That Is
Student-side deadlines cluster tightly. The shortest found is two days to choose how a case is resolved. Appeal windows run from three calendar days to ten working days, and they typically start when the notice is sent rather than when it is read.
| WHAT THE STUDENT MUST DO | TIME ALLOWED | SOURCE |
|---|---|---|
| Choose a resolution route (NC State) | two days | North Carolina State University |
| File an appeal, most common window | 10 working days | University College London |
| Respond before the case proceeds regardless (UCL) | 10 working days | University College London |
| Appeal to the Provost (CT State) | three (3) calendar days | CT State Community College |
| Accept or decline early resolution (Adelaide) | five working days | Adelaide University |
The distinction between calendar days and working days does more damage than any other technicality in these documents. A three-calendar-day appeal window that opens on a Friday is functionally a one-day window. Several procedures also start the clock from the date a notification is sent rather than received, which quietly transfers the cost of a slow inbox, a spam filter or a term-time absence onto the student. Neither choice is hidden. Both are easy to miss when you are reading the letter that says you have been accused of cheating.
There is a second asymmetry buried in these numbers. The student’s deadlines are almost all front-loaded, falling in the period when they know least about the case. UCL’s procedure is explicit that a department may proceed without the student’s input if they fail to respond within the stated window, which means the most consequential deadline arrives before most students have worked out what evidence they even have. If you are inside that window now, the fastest useful thing you can do is assemble whatever version history your documents kept and understand how track changes reads as evidence of process, because both degrade the longer you wait.
None of the nineteen procedures audited offers a student the thing the institution routinely grants itself, which is an extension for complexity. Several allow an extension on request, at someone’s discretion. That is not the same as a rule, and it is not the same as the language institutions use about their own timetables, where complexity is an accepted and unpenalised reason for a case to run long. If you plan to ask for more time, the defence checklist for a Turnitin AI flag is a better basis for the request than an apology.
4 The Only Published Answers to How Long This Takes
No institution publishes how long its cases actually run. Four procedures put a number on it in advance. The longest formally published investigation window is sixty to ninety days, and the only end-to-end cap on a whole procedure is ninety calendar days.
| PUBLISHED WINDOW | TIME STATED | SOURCE |
|---|---|---|
| Longest published investigation window (UC Berkeley) | sixty (60) to ninety (90) days | UC Berkeley |
| Only end-to-end cap on a whole procedure (Brunel) | 90 calendar days | Brunel University London |
| Charge letter after a complaint is filed (UC Berkeley) | ten (10) days | UC Berkeley |
| Allegation report to first-stage decision (Canterbury) | 35 working days | Canterbury Christ Church University |
Berkeley’s figure is the single most useful number in this audit, and not because Berkeley is slow. It is useful because it is an institution stating, in its own governing document, that concluding an investigation promptly means somewhere between two and three months. Students arriving at month two convinced that something has gone wrong are, by that standard, still inside the normal range. The wait they are experiencing as a malfunction is the published design.
Brunel’s ninety-calendar-day cap is the closest thing in the audit to an end-to-end promise, and it is worth noticing how rare that is. Almost every other institution that commits to a timetable commits to it stage by stage, which means the stages can be counted but the total cannot. A procedure with a twenty-working-day panel window, a five-working-day notification window and an unbounded gap between the flag and the start of the investigation has published a great deal without telling you when this ends.
That unbounded gap at the front is the part no document in this audit measures. Berkeley’s clock starts when the student is notified of the allegations. Canterbury’s starts from the allegation report. Nothing governs the stretch between the moment an assignment is flagged and the moment somebody decides to act on it, which is exactly the stretch students spend refreshing their email. Given how often the flag itself turns out to be contestable, whether through documented bias against non-native writers or the broader litigation now attaching to detection tools, an unmeasured delay before anyone has even reviewed the evidence is a strange place for the process to be silent.
5 Why the Queue Exists, and Why Nobody Is Counting It
Volume explains the backlog. One university reported handling over two thousand four hundred misconduct cases with roughly three quarters of them AI-related, and said it hired staff and stayed behind. No regulator in the audit publishes duration data.
| MEASURE | FIGURE | SOURCE |
|---|---|---|
| Student misconduct cases handled at LSU | 2,400 | WAFB 9 News |
| Share of those cases that were AI-related | 73% | WAFB 9 News |
| Timescale data published by the UK ombudsman | None | Office of the Independent Adjudicator |
The economics here are simple and unflattering. Running a detector across a cohort costs almost nothing and scales instantly. Adjudicating what comes out of it costs staff time and scales not at all. When one institution reports that roughly three quarters of its misconduct caseload is now AI-related, the resulting queue is not a sign that anyone is being obstructive. It is the arithmetic of a cheap referral mechanism feeding an expensive review process, and the cost of the mismatch is paid in the one currency the student cannot recover, which is time.
What makes this hard to challenge is that nobody measures it. The Office of the Independent Adjudicator, which reviews these complaints in England and Wales once internal procedures conclude, publishes a casework note on artificial intelligence and academic misconduct that sets out expectations about evidence and fairness and says nothing at all about how long providers should take. If the regulator does not collect duration data, and the institutions do not publish it, then the only people who know how long these cases run are the people waiting, and their accounts do not aggregate into anything anyone will cite.
That absence is the real finding of this audit, and it is the reason the article you are reading is a table of policy documents rather than a table of outcomes. The honest state of the evidence is that published deadlines can be compiled and compared, while actual elapsed time cannot be sourced at all. Until an institution or a regulator starts reporting it, the most defensible thing a student can say about their own wait is what their procedure promised, which is why it is worth knowing whether the rules let a detector score stand as proof in the first place, and worth keeping a defence file intact for however long this runs.
Explore every figure in this article
Methodology
This audit compiled nineteen published academic misconduct procedures from institutions in the United States, the United Kingdom, Australia and Canada, selected because their procedure documents are public and reachable without a login, and read each one for two things: the deadlines it imposes on the student and the deadlines it imposes on itself. Institutions were not sampled randomly and the set is weighted toward English-language systems, so the counts describe this collection rather than higher education generally. Every figure quoted is a stated policy limit, not an observed outcome, because no institution in the audit publishes median actual elapsed time and no regulator consulted collects it; where a procedure softened a limit with a qualifier such as normally, that limit is still counted as stated. Six figures could not be read by automated checking because the source is a PDF or the host blocks scripted clients, specifically the UC Berkeley, CT State and James Cook University documents, and each of those was confirmed by opening the document directly on 19 September 2026. All three research APIs used upstream in this project ran at full strength, so no stage of the underlying research was degraded, with one exception recorded during the recency sweep: the YouTube leg returned results unrelated to the query for one of four queries and was marked unreliable for that query, and no fact in this article rests on it.
- Sources consulted: 47
- Sources cited: 22
- Data freshness: current year: 18, last year: 3, older: 1
- Data range: 2025-01-01 to 2026-09-19
- Research date: 2026-09-19
- Update schedule: Annually, or when a procedure is revised
- Limitations: No institution publishes median actual elapsed time, so every duration figure here is a stated policy limit rather than an observed outcome. The audit covers 19 published procedures across four countries and is not a random sample. Six figures could not be machine-verified because the source is a PDF or the host blocks automated clients (UC Berkeley, CT State and James Cook University); each was confirmed by opening the document directly on 19 September 2026.
Frequently Asked Questions
How long does an AI academic misconduct case take?
There is no published median, because no institution releases one. The best available answer is what the rules promise: UC Berkeley's Code states investigations conclude generally within sixty to ninety days of notification, and Brunel caps its entire procedure at 90 calendar days. Most procedures set no end-to-end limit at all. (UC Berkeley, 2026)
Is my university allowed to take months to decide?
In most of the procedures audited, yes. Six of the nineteen set no deadline at all for the institution to reach a decision, and where a limit exists it is usually softened by the word normally. The deadlines that are firm are almost always the ones pointed at the student. (North Carolina State University, 2026)
What is the shortest deadline I will face as a student?
Two days, at NC State, to choose how the case is resolved. CT State allows three calendar days to appeal to the Provost, which a weekend largely consumes. Ten working days is the most common appeal window across the UK procedures reviewed. (North Carolina State University, 2026)
Why are these cases taking longer than they used to?
Volume. LSU reported handling more than 2,400 student misconduct cases with 73% of them AI-related, and its provost said the university hired additional staff yet remained behind on the backlog. Detection scales instantly; adjudication does not. (WAFB 9 News, 2026)
Can I complain about how long my case is taking?
In England and Wales the Office of the Independent Adjudicator reviews complaints once internal procedures conclude, but its casework note on AI cases publishes no timescale data and sets no duration standard. Delay alone is rarely the ground that succeeds; procedural unfairness is. (Office of the Independent Adjudicator, 2025)
Does the clock start at the flag or at the notification?
Almost always at notification, which is why the wait feels longer than the policy reads. Berkeley measures its sixty-to-ninety-day investigation window from when the student is notified of the allegations, not from when the assignment was flagged. Time spent before that letter is unmeasured. (UC Berkeley, 2026)
Which universities actually bind themselves to a decision date?
In this audit, the UK and Australian procedures mostly do: UCL, Brunel, Canterbury Christ Church, UEL, UWL, Adelaide and JCU all state institution-side limits. Among US procedures it is a minority, with Berkeley, Maryland, Wisconsin-Madison, Indiana, Cornell and CT State stating one. (University College London, 2026)
Sources & References
- UC Berkeley. “UC Berkeley – academic misconduct procedure.” conduct.berkeley.edu/wp-content/uploads/2026/03/Interim-Code-of-Student-Conduct-. Accessed 2026-09-19.
- North Carolina State University. “North Carolina State University – academic misconduct procedure.” studentconduct.dasa.ncsu.edu/students/procedures-overview/. Accessed 2026-09-19.
- Brunel University London. “Brunel University London – academic misconduct procedure.” students.brunel.ac.uk/documents/Policies/Academic-Misconduct-Procedure-2025.pdf. Accessed 2026-09-19.
- WAFB 9 News. “WAFB 9 News – academic misconduct procedure.” wafb.com/2026/08/20/lsu-expanded-resources-keep-up-with-ai-cheating-cases-provos. Accessed 2026-09-19.
- Office of the Independent Adjudicator. “Office of the Independent Adjudicator – academic misconduct procedure.” oiahe.org.uk/resources-and-publications/learning-from-our-casework/ai-and-academ. Accessed 2026-09-19.
- University of Maryland. “University of Maryland – academic misconduct procedure.” studentconduct.umd.edu/you/students/ai-appeal. Accessed 2026-09-19.
- University College London. “University College London – academic misconduct procedure.” ucl.ac.uk/study/current-students/academic-manual/chapters/chapter-6-student-case. Accessed 2026-09-19.
- Indiana University Bloomington. “Indiana University Bloomington – academic misconduct procedure.” studentcode.iu.edu/procedures/bloomington/discipline/academic-misconduct/index.h. Accessed 2026-09-19.
- Missouri State University. “Missouri State University – academic misconduct procedure.” missouristate.edu/Policy/Chapter3/Op3_02_AcademicIntegritySummary.htm. Accessed 2026-09-19.
- University of Wisconsin-Madison. “University of Wisconsin-Madison – academic misconduct procedure.” conduct.students.wisc.edu/academic-misconduct/. Accessed 2026-09-19.
- Arizona State University (Engineering). “Arizona State University (Engineering) – academic misconduct procedure.” students.engineering.asu.edu/policies/academic-integrity/violation-process/. Accessed 2026-09-19.
- CT State Community College. “CT State Community College – academic misconduct procedure.” ctstate.edu/images/Forms-Documents/Conduct/Academic-Misconduct-Process.pdf. Accessed 2026-09-19.
- Cornell University. “Cornell University – academic misconduct procedure.” deanoffaculty.cornell.edu/faculty-and-academic-affairs/academic-integrity/upcomi. Accessed 2026-09-19.
- University of Memphis. “University of Memphis – academic misconduct procedure.” memphis.edu/osa/pdfs/academic_misconduct_form.pdf. Accessed 2026-09-19.
- Solent University. “Solent University – academic misconduct procedure.” students.solent.ac.uk/official-documents/quality-management/academic-handbook/4l. Accessed 2026-09-19.
- Canterbury Christ Church University. “Canterbury Christ Church University – academic misconduct procedure.” canterbury.ac.uk/asset-library/policy-zone/Student-Academic-Misconduct-Procedure. Accessed 2026-09-19.
- University of East London. “University of East London – academic misconduct procedure.” uel.ac.uk/about/governance/manual-general-regulations/academic-integrity-academi. Accessed 2026-09-19.
- University of West London. “University of West London – academic misconduct procedure.” uwl.ac.uk/about-us/policies-and-regulations/academic-offences-regulations. Accessed 2026-09-19.
- Adelaide University. “Adelaide University – academic misconduct procedure.” adelaide.edu.au/about/policies/academic-integrity-policy/academic-misconduct-pro. Accessed 2026-09-19.
- James Cook University. “James Cook University – academic misconduct procedure.” jcu.edu.au/policy/academic-governance/student-experience/student-academic-miscon. Accessed 2026-09-19.
- University of Manitoba. “University of Manitoba – academic misconduct procedure.” umanitoba.ca/student-supports/academic-supports/academic-integrity/how-academic-. Accessed 2026-09-19.
Last updated:
