See what an evidence-first approach changes at your institution
Before vs. after handling academic integrity in iVake
Where you are today
Holding you back
A percentage treated as proof
A detector score lands on a desk and becomes an accusation, although it is a probability and not evidence.
Pain point
Honest students flagged
Students writing in their second or third language produce plain, careful English, which AI-writing detectors tend to flag unfairly.
Pain point
A race the detector loses
Paraphrasing tools and newer models slip past detectors, so a policy built on one grows weaker every term.
Pain point
Only the final copy is seen
The teacher judges one finished document, with no drafts or history to show how the work came together.
Pain point
- 1Final document to judge the work by
- NoneWritten AI-use rule for the assignment
- WeeksSpent arguing over a disputed score
Your institution, transformed
Everything connected
Evidence, not a score
Matching passages are shown beside their source, for a teacher to read and weigh.
Solved
Work seen as it develops
Outlines, drafts and the final copy are submitted in stages, with every version kept.
Solved
A rule for each assignment
Students see whether AI help is allowed, and how to disclose it, before they start.
Solved
A person decides
Staff decide a case after talking to the student, and the outcome is recorded with its evidence.
Solved
- 0AI-writing verdicts issued by software
- AllVersions of a submission kept
- 1Record per case, with the evidence attached
Ready to make the switch? See how it works for your institution.
Evidence a teacher can read, not a score
AI-writing detectors give probabilities, and across a whole school or university even a small error rate wrongly accuses real students every term. Those flagged most often are students writing in a second language. So iVake shows what a person can check instead: the passages that match, and what they match.
- Matching passages shown beside their source, for the teacher to judge
- No “written by AI” percentage attached to any student’s work
- Nothing added to the student record without a staff decision
Assessment design that shows the work developing
Undisclosed AI use is harder, and less worthwhile, when work is seen in stages. An assignment can ask for an outline, a draft and a final copy, keep every version, and say up front whether AI help is allowed and how to disclose it.
- Staged submissions, with every version kept and dated
- An AI-use rule on each assignment, shown to students before they start
- Room for a short viva, an in-class part or a prompt tied to class discussion
Cases decided by people, with the evidence on record
When a concern does come up, it should rest on what a person can defend: the version history, the matching passages and a conversation with the student. The decision, the evidence and who made the call are kept on the record.
- Cases recorded with the evidence, the outcome and the staff member who decided
- Case records seen only by the staff your policy names
- A governed study tutor in the app, so students have less reason to use outside chatbots
Academic integrity features
Everything you need to handle academic integrity fairly
Matching passages
Text matched against course material and other submissions.
No AI-writing score
No probability verdict attached to a student’s work.
Staged submissions
Outline, draft and final copy, each dated.
Version history
Every version of a submission kept.
AI-use rule per assignment
Allowed, limited or not allowed, with how to disclose it.
Approved study help
A tutor that answers from the course material.
Human decisions
Cases decided by staff after talking to the student.
Restricted & logged
Case records seen only by named staff, every access logged.
Frequently asked questions
Get answers to common questions from academic heads and integrity committees
Does iVake detect AI-written work?
No, and that is a deliberate choice. Detector scores are probabilities, they are often wrong, and they flag students writing in a second language more often. We would rather give teachers evidence they can read than a number that cannot carry the weight of a disciplinary decision.
What does iVake show instead?
Passages in a submission that match course material or other submissions, shown beside their source, and the full version history of the work. A teacher reads both and decides whether there is anything to discuss.
Should we still use a plagiarism checking service?
Many universities do, especially for dissertations and research work under UGC’s regulations on plagiarism. Treat a matching report as the start of a conversation, check the matches by hand, and never act on a percentage alone.
How does iVake help with academic integrity, if not by detection?
Through the way work is set and seen: staged submissions with version history, a clear AI-use rule on each assignment, a study tutor students are allowed to use, and a recorded decision by a person when a concern comes up.
A vendor showed us a very accurate detector. Why not rely on it?
Accuracy figures come from test sets, not from your students. Even a small false-positive rate across thousands of submissions wrongly accuses real students each term, and new models and paraphrasing tools keep changing what a detector sees.
Where should we start on an AI-use policy?
Decide, for each type of assignment, whether AI help is allowed, limited or not allowed, and how students disclose it. Put it in the handbook, show it on each assignment, and agree who decides a case and how a student can appeal.
Who can see an integrity case?
Only the staff your policy names. Case records are restricted in the same way as discipline and welfare notes, and every read and change is written to the audit trail.
Cloud, hosted in your region
REST API & webhooks
SSO: Google, Microsoft & SAML
Encrypted, backed up daily- Built for the DPDP Act
- Full security details
Built for your institution type
See how this feature works for your specific institution
UniversitiesCases decided by panels, with evidence kept
Degree collegesStaged assignments and AI-use rules per course
CBSE, ICSE & state-board schoolsProject work seen in stages, not just at the end
ITIs, polytechnics & skill centresPractical reports compared within the batch
Residential & international schoolsCoursework rules on AI use, shown per assignment
Professional collegesProject reports submitted as outline, draft and final
Integrations, comparisons & tools
Extend academic integrity with integrations, see how iVake compares, and run the numbers
Related features
Grading Assistant
Feedback drafted against your rubric, with the teacher setting every grade.
Student Tutor
Answers from your own course material, with a teacher to hand over to.
Assignment
Assignments set, submitted and returned on the student record.
Assignment Annotation
Comments marked directly on a student’s submitted work.
Quiz Anti Cheating
Controls that help keep an online quiz fair to every student.
Thesis Management
Research proposals, supervision and submissions on the scholar’s record.
More in AI tools
Explore other features in this category
Ready to handle academic integrity fairly?
See staged submissions, matching passages and recorded decisions in iVake, with a person deciding every case.
Try it free for 30 days. No credit card required.