If your programming course runs on Canvas, you have probably wondered whether uploading a .py or .java file lets the platform spot AI help. The short answer: Canvas itself is a learning management system, not a detector. What your instructor sees depends on which extra tools your school has connected to it, and the most common of those, Turnitin, says its AI writing detection is not built for code. Everything below about Canvas and Turnitin comes from Instructure's and Turnitin's own documentation, which we read in October 2026. Settings differ between schools, so check the linked pages and your own course for the details that apply to you.
What Canvas itself does with your submission
Canvas collects your file or text entry, stores it with a timestamp, and shows it to your instructor in tools like SpeedGrader and the Gradebook. That is the core job. In the documentation we read, Instructure does not describe a built-in feature that analyses a submission and estimates whether it was written by a person or by a language model.
What Instructure does document is a way for other companies to plug checking tools into Canvas. Its Plagiarism Detection Platform page describes a standard way for LTI tool providers to integrate plagiarism detection tools with Canvas, including originality reports that the tool creates and attaches to a submission. The same page now strongly recommends a newer route called Document Processors, based on the 1EdTech LTI Asset Processor standard, and says the API documentation has moved to Instructure's Developer Documentation Portal. Either way, the design is the same: Canvas passes the submission to an outside tool, and the outside tool decides what to report.
What Turnitin adds when a school connects it
Turnitin is the tool most students meet inside Canvas. Turnitin's Canvas Plagiarism Framework FAQ explains that this integration lets instructors enable Turnitin on a Canvas assignment, and that it was designed to provide a Similarity Report and spelling and grammar suggestions. The similarity percentage appears in the Canvas Gradebook, and the full report opens from SpeedGrader. Instructors choose when students can see the report: immediately, after grading, after the due date, or never.
A Similarity Report compares your submission with other text, such as web pages and earlier submissions. It is a matching tool. It answers "does this look like something else we have seen?", not "was this generated by AI?".
Turnitin also sells AI writing detection, but its own AI writing detection FAQ is direct about code. It says the model removes text that appears to be code before scoring, that it looks only for prose sentences in long-form writing, and that it does not reliably detect AI-generated text in the form of code. We cover this in more depth in does Turnitin detect AI code.
Other tools your school might connect
Because Canvas is built to accept external tools, your course could be using a different plagiarism or AI detection vendor, or none at all. Each vendor makes its own claims, and they are not Canvas features. If you want to know what is checking your work, the honest answer is to look at your assignment settings, ask your instructor, or read your school's Canvas help pages. The tool's own documentation, not a blog post about Canvas, is what tells you what it can and cannot see.
Why code is hard for AI detectors in general
Even a tool built to look for AI writing has a hard time with programs. In a 2024 study presented at ICSE, Pan and colleagues ran five general AI-content detectors on thousands of Python solutions, some written by people and some by ChatGPT. They concluded that the existing detectors performed poorly at telling human-written code from AI-generated code. Several tools labelled almost everything as human, and at least one labelled most of the human-written code as AI. Those were older, text-focused detectors, so the result does not describe every product on the market, but it shows why a score on code deserves caution. Our guide on how accurate AI code detectors are goes through more of the research.
What your instructor can still notice
- Your history in the course. Earlier submissions are all in Canvas. A sudden change in style or level is something a person can see without any tool.
- Similarity between classmates. Some programming courses run code similarity tools outside Canvas. See does MOSS detect AI code for how those work.
- Conversation. Being asked to explain a function, trace a loop or make a small change is common, and it is the check that matters most.
What this means for you
- Read your course policy. Some courses allow AI help with disclosure, others ban it. Canvas will not tell you which; your syllabus will.
- Keep your process. Commits, drafts and notes are the clearest evidence of how you worked if anyone asks.
- Do not read too much into a number. A similarity percentage on code is about overlap, and an AI score on code is, by Turnitin's own account, not reliable.
Checking how your code reads
Canvas will not show you how your code comes across, but you can look for yourself. Paste your file into the detector at the top of this page and read the reasons it gives, not just the score. If it points to habits like a docstring on every tiny helper or an example block the assignment never asked for, you will know what a grader might ask about, and you can be ready to explain the choices you made. If a tool has already flagged your work, our guide my code was flagged as AI walks through what to do next.
Frequently asked questions
Does Canvas have a built-in AI detector?
We found no Instructure documentation describing one. Instructure's developer documentation treats plagiarism detection as something third-party tools plug into Canvas, and those tools create the reports instructors see. If your course shows a similarity or AI score, it comes from a tool your school added, not from Canvas itself.
My Canvas assignment shows a Turnitin percentage. Does that include my code?
The Turnitin percentage you see in the Gradebook or SpeedGrader is usually the Similarity Report, which shows matching text. Turnitin's separate AI writing detection removes text that looks like code before scoring, and Turnitin's FAQ says the model does not reliably detect AI-generated code.
Can my instructor see that I used ChatGPT from Canvas alone?
Canvas stores what you submit, and your instructor reads it in SpeedGrader. Nothing in the documentation we read describes Canvas labelling a submission as AI-written. Instructors can still notice code that does not match your earlier work or ask you to explain it.
What if my school uses a different plugin, not Turnitin?
Then the answer depends on that vendor. Canvas lets institutions connect several kinds of plagiarism tools, and each one has its own documentation and limits. Ask your instructor or check your school's Canvas help pages to find out which tool is enabled for your course.
Is it safe to rely on a 0% score?
No. A low or missing score on a code submission often just means the code was not analysed, or was not similar to anything in the comparison set. It is not a statement that your code reads as human. Follow your course's AI policy and keep your drafts and commit history.
Try the detector
Have a snippet you are unsure about? Paste it into the free AI Code Detector to get a verdict, an AI probability and the reasons behind it. No sign-up, and your code is not stored.