How to Adjust the AI Content Detection Sensitivity in Copyleaks

Copyleaks AI content detection sensitivity has three levels.

Level 1: Detecting content copied directly from a large language model (LLM), like ChatGPT or Gemini, without edits.

Level 2: Detecting content from an LLM with minor changes, like tense adjustments or added words.

Level 3: Detecting content from an LLM that has been heavily modified using tools or manual edits.

By default in our Canvas instances, Copyleaks AI content detection sensitivity is set to level 2. If you'd like to update that setting (to, for instance, lower the sensitivity level), this guide provides directions for how to do so in Canvas.

Note: You can only edit the AI Content Detection Sensitivity before students submit the assignment. Once students submit, you cannot edit this setting.

Canvas Directions:

Navigate to SPC's Canvas instance, navigate to your course, and then select Modules. Select the + inside the module to add a new assignment to the module. If you've already created the assignment, skip ahead to step 3.
Name the assignment. Then select the assignment and select Edit.
Edit the assignment details (points, attempts, etc., and submissions type should be online via file upload).
Under Plagiarism Review, select Copyleaks Plagiarism Checker from the dropdown menu.
From the initial Copyleaks menu, select Advanced Settings.
In the advanced settings menu, select either course settings (to change settings for all Copyleaks-enabled assignments in the course) or assignment settings (to change settings for just this assignment) (1). Then, in the scan settings tab (2), ensure that AI content detection is toggled on (3), and then you'll have the opportunity to choose how sensitive you'd like the AI detection to be (4) before clicking save (5).