The short answer
No. Canvas has no built-in AI detector. AI scores inside Canvas come from a tool the school connects, such as Turnitin, and only when it is licensed, switched on and attached to the assignment. As of October 2026, an assignment with no connected tool gets no AI check from Canvas itself.
Does Canvas have a built-in AI detector?
No. Canvas is a learning management system (LMS), and the AI scores people see inside it come from outside tools that a school connects. Instructure's developer documentation calls the LTI 1.3 Asset Processor, which Canvas names the Document Processor, "the new standard for conducting plagiarism checks with external tools". The design is simple: Canvas passes the submitted file to a connected tool and shows the report that comes back.
Detector vendors describe it the same way. GPTZero's Canvas integration page says Canvas "does not currently include AI detection" and that a school adds it by connecting GPTZero. That is a vendor claim, but it fits Turnitin's AI writing FAQ, which says no AI score is embedded in the LMS interface and users have to open the report itself.
How do AI checkers plug into Canvas?
There are two routes today. The older one is the Canvas Plagiarism Framework (CPF): the instructor sets the submission type to Online with Text Entry or File Uploads, then picks the school's Turnitin integration from the Plagiarism Review menu, as Turnitin's setup guide explains. The newer one is the Document Processor: the instructor scrolls to the bottom of the assignment settings and selects Add Document Processing App, as Turnitin's Canvas workflow guide shows.
Instructure's documentation describes CPF as deprecated. In July 2026, Turnitin wrote that Canvas will retire CPF over time and that Similarity and AI writing detection "continue to run directly on native Canvas assignments as they did with CPF". Schools can also use Turnitin's own assignment type, which since July 2026 can be marked in SpeedGrader.
What AI detector does Canvas use?
Canvas does not choose one. Each school chooses and pays for its own tool, so two courses on the same Canvas site can have different checks or none at all. Turnitin, Copyleaks and GPTZero each publish Canvas setup material, and the table shows how they connect and how they scored in Plagino's own test.
| Detector | How it reaches Canvas (vendor information) | AI documents caught | Human documents wrongly flagged |
|---|---|---|---|
| Turnitin | Through CPF or the Document Processor. AI writing detection needs a Turnitin Originality license. | 98.2% | 1.4% |
| Copyleaks | Copyleaks describes a Canvas plugin with AI detection and publishes a Canvas Asset Processor (LTI 1.3) setup article. | 91.3% | 2.2% |
| GPTZero | GPTZero describes a one-time integration per institution, with results shown in SpeedGrader or assignment views. | 88.9% | 5.7% |
Which of these your course uses is a school decision, not a Canvas setting you can see from the student side. You can read each vendor's own description on Copyleaks' Canvas page, its Canvas Asset Processor setup article and GPTZero's Canvas page. For tool-by-tool comparisons, see Plagino vs GPTZero and Plagino vs Copyleaks, or the full best AI detectors round-up.
Can Canvas detect ChatGPT, Claude or Gemini?
Canvas on its own does not score writing for AI, but a connected detector can. If the school has Turnitin Originality with AI writing detection switched on, the answer is yes for the models Turnitin lists. As of October 2026, Turnitin's AI writing FAQ says its English model can detect GPT-4o and the GPT-5 family up to GPT-5.4, Gemini models up to 3.1 Pro preview, Claude models from Claude 3 Haiku through Sonnet 4.6 and Opus 4.5, plus Llama, Mistral, DeepSeek, Nova and Grok models, and tools built on them.
That list is Turnitin's own claim, and a model that is not on it is not promised. The same FAQ says Turnitin only processes a submission for AI writing if the institution has the feature enabled, and that an admin can turn it off. In Plagino's March 2026 benchmark, whose AI half came from GPT-4o, Claude and Gemini, Turnitin caught 98.2% of 500 AI-written documents and wrongly flagged 1.4% of 500 human-written ones. Our guide on whether Turnitin detects AI covers the rest of its rules.
How does Canvas Turnitin AI detection work for instructors?
The instructor opens the submission in SpeedGrader and finds the report in the right-hand panel. Turnitin's guide says to select View Report under Document processor reports for the Turnitin Similarity and AI writing report, if AI writing is enabled for the institution. Inside the report, the AI Writing tab holds the overall percentage of text detected as AI.
The AI percentage and the similarity score are separate numbers. Turnitin says they are independent and do not influence each other, and AI highlights do not appear in the Similarity Report. To get an AI score at all, a file needs at least 300 words of prose and must be a .docx, .pdf, .txt or .rtf file. Our explainer on the Turnitin similarity score covers the matching side.
How can you tell if a Canvas assignment uses AI detection?
Canvas does not show students a simple on or off label for AI checks, so work through these steps in order. None of them is perfect, which is why the last step is to ask.
- 1
Read the syllabus and the assignment page. Look for an AI policy or a mention of Turnitin or another checker. If the course is silent, note that and ask later. A policy also tells you what counts as allowed AI use.
- 2
Check the submission type. The documented Canvas setups attach to assignments that accept a text entry or a file upload. Quizzes and other submission types are not part of the Document Processor setup that Instructure describes.
- 3
Watch for a license agreement prompt. Instructure's documentation says Canvas can show a connected tool's end user license agreement when a student opens an assignment with a document processing app. One university reports that students on the Canvas mobile app may not see it, so use a web browser if the prompt is missing.
- 4
Look in Grades or Assignments after submitting. If the instructor allowed it, students can open their report from Grades or Assignments. That link may cover similarity only, because Turnitin says the AI indicator and report are not visible to students unless an instructor shares a PDF.
- 5
Ask your instructor directly. Ask which tool is attached, whether AI detection is switched on, and whether you will see the result. A direct answer beats guessing from the interface.
Does Canvas check discussion posts or quizzes for AI?
Usually not through the same route. Instructure's Document Processor attaches to assignments that accept text entry or a file upload, so a discussion post or a quiz answer is outside that setup. In a December 2023 Instructure Community thread, an answer said there was no officially supported way to check discussion posts for plagiarism or AI. That answer is dated, so ask your school whether anything has changed.
Third-party discussion tools may add their own checks. The University of Illinois Springfield says the Turnitin check inside its Harmonize discussions does not include generative AI detection and is not as robust as the full checker. An instructor who suspects a post can still paste it into a separate tool by hand, which is where the one-document-per-student rule below matters.
Why should each student's work be checked in its own file?
Because the detector reads text in blocks. The same university warns that if several students' answers sit in one document, a human-written answer sandwiched between AI-written ones is likely to be falsely flagged. Putting each response in its own document is the way to assess each one as fairly as possible. Teachers can read more in our guide to using an AI detector as a teacher.
Can students see the AI score in Canvas?
Not by default. Turnitin's FAQ states that the AI writing indicator and report "are not visible to students", although an instructor can download the AI report as a PDF and share it. The similarity side works differently. With CPF, instructors choose whether students see the Similarity Report immediately, after grading, after the due date or never.
With the Document Processor, the choice is made once. Turnitin's guide says the setting that lets students view the report after submitting is locked after the assignment is saved, and Instructure says students only see a report the tool marks as visible to them. If you want to see what a detector would say about your own draft before you submit, run it through the Plagino AI detector.
How long does Turnitin take on Canvas?
Usually minutes, with longer waits for resubmissions. Turnitin's own guidance gives the figures in the table, and all of them are Turnitin statements as of October 2026. If a report looks stuck, the likely cause is the file, not Canvas.
| Situation | What Turnitin says |
|---|---|
| A newly uploaded file | Can take up to 10 to 15 minutes to process and produce a Similarity Report. |
| The AI writing report | Generating the results can take several minutes, and the indicator shows a loading state until then. |
| A resubmitted file | May take up to 24 hours to produce a new Similarity Report. On classic Standard Assignments the first three resubmissions generate straight away, then a 24-hour wait applies. |
| First submission to a new CPF assignment | Wait 90 seconds after creating the assignment so Canvas can tell Turnitin it exists. |
| A report that never appears | In CPF, the instructor can select Resubmit to Turnitin in SpeedGrader to force the report. |
| Files that often fail | Dense vector images, an image-only first page, password-protected files and unusual file types. Converting to PDF from Word or Google Docs is the suggested fix. |
The 10 to 15 minute figure is for the Similarity Report. See Turnitin's report loading guide and resubmission rules if yours takes longer. Turnitin's AI writing guide describes the AI report as a separate loading state.
Does Moodle have AI detection?
Not as a built-in feature that we could find in its documentation. Moodle's AI tools page describes an AI subsystem for generating text and images, summarising and explaining course content through providers such as OpenAI, Azure AI and Ollama. That is about helping users write, not about detecting AI writing.
Detection comes from plugins. Turnitin offers a Moodle plugin, and its FAQ says the AI indicator is available through Moodle, Blackboard and Canvas. Originality.ai says its Moodle plugin scans assignments, quizzes and forum posts for AI and plagiarism, which is a vendor claim. In Plagino's March 2026 benchmark, Originality.ai caught 95.6% of AI documents and wrongly flagged 4.1% of human ones.
Do Brightspace, Blackboard and Gradescope have AI detection?
The pattern repeats: the platform manages coursework, and an add-on does the detecting. Each platform below differs in the details.
Brightspace
D2L's Turnitin article for Brightspace describes Similarity Report and Online Grading features, with the instructor enabling Turnitin per assignment under Evaluation & Feedback. Villanova University's Brightspace guide says instructors can access AI writing detection information where available. A D2L community answer says the AI report is instructor-only unless the instructor shares a PDF.
Blackboard and SafeAssign
Blackboard's own SafeAssign documentation describes SafeAssign as a text-matching algorithm that compares papers against sources, and notes that Blackboard has a partnership with Turnitin, which it points to for languages other than English. The page does not describe an AI-writing detector, so a SafeAssign score on its own is a matching result, not an AI score.
Gradescope
Turnitin says it has no current plans to add AI detection to Gradescope. Its FAQ explains that Gradescope's main use is handwritten work, while Turnitin's AI detection focuses on typed text, and that Turnitin is not pursuing ChatGPT code detection at this time.
How do the major learning platforms compare on AI detection?
This table pulls the findings above into one view. It reflects vendor and school documentation as of October 2026, and a school's own setup can differ.
| Platform | Built-in AI writing detector? | How AI detection is added | Student visibility |
|---|---|---|---|
| Canvas | None documented. Instructure points plagiarism tools to the Document Processor. | Turnitin through CPF or the Document Processor. Copyleaks and GPTZero describe Canvas integrations. | Turnitin AI report hidden from students by default. Instructor can share a PDF. |
| Moodle | None found. Moodle's AI tools generate and summarise text. | Turnitin plugin. Originality.ai says it has a Moodle plugin. | Turnitin AI report hidden from students by default. |
| Brightspace | None documented in D2L's Turnitin article. | Turnitin integration, with AI information where licensed. | AI report is instructor-only per a D2L community answer. |
| Blackboard | SafeAssign is described as text matching. | Turnitin is mentioned as a partner in Blackboard's SafeAssign page. | Not covered in the sources we checked. |
| Gradescope | None. Turnitin has no current plans to add one. | Not offered by Turnitin for Gradescope. | Not applicable. |
How reliable is AI detection inside an LMS?
The LMS does not change how accurate the detector is. Turnitin says its percentage "should not be used as the sole basis for adverse actions against a student" and that scores from 1% to 19% show only as an asterisk because false positives are more common there. For short documents of a few hundred words, the same FAQ says the prediction is mostly all or nothing. The University of Illinois Springfield calls these reports a starting point for a conversation, not proof either way.
Independent evidence agrees that detectors can be wrong. A 2023 Stanford-led study found that several detectors misclassified more than half of TOEFL essays by non-native English speakers as AI-generated. In Plagino's March 2026 benchmark, Plagino wrongly flagged 0.8% of human documents and Turnitin 1.4%, while every other tool flagged more. Read our explainer on why AI detectors produce false positives before treating any score as proof.
What should students and teachers do?
Students should keep their process. Save outlines, notes and dated drafts in a tool with version history, and ask the instructor which AI uses the course allows before starting. If a flag is ever raised, that record carries more weight than a second detector's score. Our guide on what to do if you are falsely accused of using AI explains how to respond.
You can pre-check a draft with a separate tool such as the Plagino AI detector, but it will not reproduce your school's Turnitin result. Our guide to a Turnitin AI checker alternative explains what a pre-check can and cannot tell you. Teachers should state their AI policy in the syllabus, tell students which tool is attached, and treat any score as one input among several.
Frequently asked questions
Does Canvas detect AI writing by itself?
No. Canvas has no AI detector of its own. Any AI score inside Canvas comes from a connected tool, such as Turnitin, that the school has licensed and attached to the assignment.
What AI detector does Canvas use?
It depends on the school. Turnitin publishes Canvas setup guides, and Copyleaks and GPTZero say they offer Canvas integrations too. A course with no connected tool gets no automatic AI check from Canvas.
Does Canvas automatically include Turnitin?
No. The school's administrator configures Turnitin with Canvas, and the instructor then attaches it to each assignment. Turnitin's AI writing detection also needs a separate Originality license.
Can Canvas detect Claude or Gemini?
Only through a connected detector. Turnitin's FAQ lists Claude and Gemini models among those its English model targets as of October 2026. Canvas itself does not identify which AI tool wrote a text.
Can Canvas detect AI without Turnitin?
Only if the school connects another detector. Copyleaks and GPTZero both describe Canvas integrations. If no tool is attached, an instructor would have to check suspect text by hand in a separate tool.
Does Canvas check discussion posts for AI?
Not through the standard Document Processor route, which covers text entry and file upload assignments. One university says the Turnitin check in its discussion tool does not include generative AI detection. Ask your school if it uses a different setup.
Can students see the Turnitin AI report in Canvas?
Not by default. Turnitin says the AI indicator and report are not visible to students, but an instructor can download the AI report as a PDF and share it. Student access to the similarity report is a separate setting.
How long does Turnitin take on Canvas?
A new file usually takes up to 10 to 15 minutes to produce a Similarity Report, according to Turnitin. A resubmitted file can take up to 24 hours. The AI writing report can take several minutes to load.
What happens if my school turns Turnitin's AI feature off?
Turnitin will not process submissions for AI writing detection. Its FAQ says admins can enable or disable the feature, and that past submissions are not scored retroactively. Similarity checking is a separate feature.
Does Moodle have AI detection?
Not as a built-in feature that we found in Moodle's documentation. Its AI tools help users generate and summarise text. Schools add detection with plugins, such as Turnitin's or Originality.ai's.
Does Gradescope have AI detection?
No. Turnitin says it has no current plans to add AI writing detection to Gradescope, because Gradescope is mostly used for handwritten work and Turnitin's detector focuses on typed text.
Is a Canvas AI score proof of cheating?
No. Turnitin says its AI percentage should not be the sole basis for action against a student and that it does not decide whether misconduct happened. That decision rests with people applying the school's academic integrity policy.
Can I check my paper for AI before I submit it in Canvas?
Yes, with a separate tool. The Plagino AI detector lets you scan a draft first. It will not match your school's Turnitin score, so treat it as a check on your own writing, not a prediction of a grade or a flag.
