Does Canvas detect AI? How LMS detection actually works

Canvas is a learning management system: it stores assignments, collects submissions and holds grades. Its assignment settings include a "Plagiarism Review" dropdown, and Turnitin's own guide for Canvas describes choosing Turnitin from that menu as the step that connects a submission to Turnitin's checks (Turnitin, creating an assignment with Canvas Plagiarism Framework, read 9 November 2026). Without a tool selected there, a Canvas submission is stored and graded and nothing scores it for AI.

Instructure, the company behind Canvas, announced in July 2025 a partnership with OpenAI to build assignments in which students work with a language model inside Canvas and teachers see the conversation (Instructure press release, 23 July 2025, read 9 November 2026). That is AI inside Canvas for teaching, with the instructor given visibility of the chat; it is not a detector run over a finished essay.

Turnitin answers the question directly in its FAQ: users "will be able to see the indicator and the report via the LMS they're using", but "there is no AI writing indicator or score embedded directly in the LMS user interface", so an instructor opens the Similarity Report to see the AI percentage (Turnitin, AI writing detection FAQs, read 9 November 2026). The feature is licensed through Turnitin Originality or as an iThenticate add-on, and Turnitin states it processes a submission for AI only when the institution has the feature enabled, read 9 November 2026.

Blackboard's documentation describes the same arrangement. Anthology's help page for Blackboard Learn says the Turnitin integration is "on by default" as an institution-level setting that can be switched off, and that student submissions are "matched against the Turnitin database" (Anthology, Blackboard Learn Turnitin help, read 9 November 2026). Anthology's page on AI plagiarism offers no detector of its own; it recommends assessment designs that are harder to answer with a chatbot, read 9 November 2026. Moodle's plugin directory lists the Turnitin plagiarism plugin, "developed and maintained by Turnitin" and requiring a licence, alongside Copyleaks and Originality.ai plugins that add AI detection (Moodle plugins directory, Turnitin plagiarism plugin, read 9 November 2026).

GPTZero, one of the five detectors in this site's run, lists Canvas and Google Classroom among its integrations, and sells institutions an enterprise plan with LMS rollout support (GPTZero for educators, read 9 November 2026). So "Turnitin" is the common case, not the only one: a school can connect a different detector, or none.

Turnitin's FAQ describes four states for the indicator an instructor sees after opening the report, read 9 November 2026. Blue with a percentage means the document processed and the number is the share of qualifying prose sentences the model labelled AI, not the share of the whole file, because lists, headings, references, tables and code are stripped first. An asterisk means AI was detected below 20 percent and no number is shown. Grey dashes mean the file could not be processed, for example under 300 words of prose. An exclamation mark is a processing error.

Turnitin also states what the number is for. It "does not make a determination of misconduct"; the percentage "should not be used as the sole basis for action or a definitive grading measure by instructors", read 9 November 2026. The accuracy claim Turnitin publishes is a false positive rate under 1 percent for documents with more than 20 percent AI writing, tested on over 700,000 pre-ChatGPT papers before each release, read 9 November 2026. The guide to how Turnitin detects AI covers the scoring in full, and the guide to whether AI detectors work sets vendor claims against independent studies.

No detector score should be the only evidence in a misconduct case.

The question answers itself once the plumbing is clear: Canvas detects nothing, so there is no "Canvas detector" to write for. What checks the work is Turnitin or whichever tool the institution connected, and Turnitin states that its indicator covers text modified by "AI paraphraser or bypasser (also called humanizers) tools", read 9 November 2026. Turnitin's detector is not available to the public, so this site relies on Turnitin's published statements and on peer-reviewed studies and states no Turnitin result for any tool.

The Turnitin evidence page lists all twelve tools on the same plan fields with no pass or fail promised for any of them, and the best-for hub holds the other use-case pages. Anyone submitting through an LMS should assume the strictest tool their institution licenses is connected, read the syllabus, and disclose AI help where the course allows it.