Q: What is Copyleaks?
A: Copyleaks is a comprehensive AI content and plagiarism detector designed for both academic and business use. It identifies text generated by models like ChatGPT, detects paraphrased content, and supports over 100 languages, integrating directly into Canvas.
Q: Is AI detection the same as plagiarism detection?
A: No. AI detection looks for text generated by AI systems (for example, large language models and AI paraphrasers). Plagiarism detection looks for matches to existing text in external and internal sources. AI-generated text can sometimes overlap with plagiarism, but the two checks are conceptually distinct.
Q: What similarity or AI threshold should I use?
A: There is no universal. You need to set your own threshold(s), consistent with your course goals, and clearly communicate them to students.
Q: Which browser works best with Copyleaks?
A: Google Chrome is the recommended browser but you can work with Copyleaks through all major browsers. Here is a link to an article that has troubleshooting steps you can try if your browser isn't working properly. Copyleaks browser compatibility issues.Links to an external site.
Q: What’s the best way to handle flagged AI reports?
A: Treat detection formatively rather than purely punitively. Verify that the correct file was submitted, review the report in detail, consider drafts and process notes, and meet with the student to discuss their writing process. Use detection scores as one piece of evidence, not sole proof of misconduct, and aim to support learning and integrity.
Q: What do the different colors on the Canvas assignment mean?
A: Canvas Similarity Score Legend
Once an assignment has been scanned for plagiarism on Canvas, a color will appear next to the report. This is how to interpret the colors you see:
Blue: No matching text
Green: 1-24%
Yellow: 25-49%
Orange: 50-74%
Red: 75-100%
Visit this link for additional detailsLinks to an external site.
Q: How does Copyleaks handle multiple submissions, and can students resubmit after seeing their score?
A: If multiple submissions are allowed in Canvas, students can resubmit after seeing their score/report. Copyleaks focuses on the latest submission for that assignment; earlier submissions are not compared against the latest for plagiarism or AI. Encourage students to review their report, revise their work, and resubmit within your allowed attempts and deadlines.
Q: Does Canvas show color-coded flags for plagiarism?
A: Yes. Canvas displays color-coded flags (typically red, yellow, or green) in the gradebook to indicate potential plagiarism. These flags relate to plagiarism/similarity, not AI detection.
Q: Is the first sign of plagiarism usually the color-coded flag?
A: Yes. For most faculty, the color-coded flag in the gradebook is the first indicator that a submission may need closer review. The flag is a prompt to open the Copyleaks report, not a final determination.
Q: Will paraphrased AI text be detected?
A: Often yes, especially at higher sensitivity levels. If students take AI-generated text and run it through paraphrasers, Copyleaks can still detect AI patterns. At Level 2 sensitivity or above, paraphrased AI content is frequently still flagged. Simply rephrasing AI output is not a reliable way to avoid detection.
Q: How does using AI tools like Grammarly or other proofreaders affect grading and detection?
A: AI proofreaders and writing assistants (such as Grammarly and similar tools) may cause work to be flagged as AI-generated because they actively rewrite or heavily modify text. These tools leave AI “fingerprints” in the writing. If your course requires students to disclose AI use, their AI score will likely be higher when they use these tools.
Q: Why is there a built-in grammar checker, and how does it work?
A: The built-in grammar checker is intended to reduce reliance on external AI tools and minimize confusion over AI scores. It provides basic error highlighting so students can improve their writing without a full AI rewrite. Students must manually correct the errors; the tool does not automatically rewrite the text for them.
Q: Will using the built-in grammar checker flag a student’s work as AI-generated?
A: No. Corrections made using the built-in grammar checker do not cause the work to be flagged as AI-generated. In contrast, papers processed through external AI tools like Grammarly are more likely to be flagged as AI-generated.
Q: Will students be “punished” if their writing naturally improves over time?
A: No. Copyleaks is designed to distinguish between normal growth in writing ability and AI-generated text. Natural improvement across the semester is not expected to be flagged as AI. Manual grammar correction and revision are fine.
Q: Can AI “source checker” tools trigger AI detection?
A: Yes, very likely. AI-based “source checker” or “improvement” tools often rewrite or heavily alter content. Copyleaks looks at syntax, structure, and other signals; the resulting text is usually identifiable as AI-generated and may be flagged.
Q: Does Copyleaks just look for highlighted “AI phrases”?
A: No. The highlighted “AI phrases” are only one small signal and are provided as a helpful starting point. Copyleaks uses many additional signals—statistical, structural, and model-based—when determining AI usage. For a full picture, instructors should consult the AI Content tab and AI Overview in the report.
Q: What if a student trains an AI to “write like me”?
A: In theory, an AI model could be fine-tuned on a large volume of a student’s writing, but in practice students usually provide only a few essays, which is far from enough for true style mimicry. Tests have shown that such attempts still produce text that Copyleaks flags as AI-generated. This is not a realistic or reliable way to avoid AI detection.
Q: Is plagiarism detection or AI detection Copyleaks’ strongest feature?
A: Plagiarism checking is mature and well-established, while AI detection is the most rapidly evolving feature and receives frequent model updates. Both are robust, but AI detection is the area of fastest ongoing development.
Q: Can I restrict which file types students can submit?
A: Yes. In Canvas assignment settings, choose a submission type that allows file uploads, then use the “allowed file types” setting to restrict formats (such as DOCX or PDF). Copyleaks will only scan the file types you permit. Be sure to communicate accepted formats to students.
Q: Does it matter if students submit a Word document or copy/paste text?
A: The system can process both, but formatting can affect how content is interpreted. Best practice is to ask students to upload the original formatted file (for example, export a Google Doc to Word or PDF). Copying and pasting from a PDF or other sources can alter layout and may affect detection results.
Q: Are these tools relevant for all assignments, like chemistry labs?
A: Not necessarily. For heavily data-focused assignments (such as some lab reports or problem sets), you may not need every Copyleaks feature. You can disable or ignore tools that are not relevant to a specific assignment type.
Q: How is the plagiarism (matching) score calculated?
A: Copyleaks scans the submission while excluding any sections you have configured to omit (such as references or templates). It then compares the remaining text to internet sources and internal databases, calculates what percentage of that text matches existing sources, and displays that as the similarity/matching score.
Q: Will citations in parentheses be flagged?
A: They can be, unless you configure omissions properly. To prevent this, open the omission/exclude settings for the assignment, specify that references, citations, and similar sections should be excluded, and Copyleaks will skip those sections in the plagiarism scan. This helps ensure standard citation formats don’t inflate similarity scores.
Q: How do I handle APA templates or style guides?
A: Upload commonly used templates as “Exclude Templates.” Take your APA or other style template (or any repeated boilerplate students will reuse), upload it to the Exclude Template section for the assignment, and Copyleaks will automatically omit any matching template text from its plagiarism check.
Q: What if students use my template in another professor’s class?
A: In another course, the same template text could be flagged as plagiarism unless that instructor also excludes it. The other instructor can upload the template to their Exclude Template settings. Each instructor must upload it individually for their own assignments.
Q: Do I need to enable the plagiarism detector for each assignment?
A: Yes. You need to turn on Copyleaks for each course or assignment where you want detection.
Q: Does Copyleaks provide a grade for assignments?
A: No. Copyleaks does not assign or suggest grades. Instructors can manually enter grades within the Copyleaks report interface, and those grades can sync back to the Canvas gradebook. Matching and AI scores are for diagnostic purposes, not automatic grading.
Q: Are student analytics available for all courses or just one?
A: Typically, instructors have analytics only for the courses they teach.
Q: Should AI use be addressed in the syllabus?
A: Yes. Include a clear AI policy in your syllabus that specifies when and how AI tools may be used, what must be disclosed or cited, and the consequences for unauthorized AI use. Sample policy language may be available from your department or institution.
Q: Is there a Copyleaks mobile app?
A: No. Copyleaks does not have its own dedicated mobile app. The Canvas Instructor app can provide some access to Copyleaks-related content, but Copyleaks is primarily designed for desktop use. Mobile formatting and UI may be suboptimal, and some users report issues submitting or reviewing reports on phones. Using a desktop or laptop is recommended.
Q: Which browser works best with Copyleaks?
A: Google Chrome is the recommended browser. Many reported issues (such as “bad page requests” or access problems) are related to browser compatibility. If you encounter problems, switch to Chrome first and then follow any additional troubleshooting steps from your institution or Copyleaks support.
Q: Can Copyleaks detect text from newer AI models?
A: Copyleaks regularly updates its AI detection models to better handle text from newer AI systems. Accuracy for detecting content from newer models is expected to improve as these updates are deployed.
Q: What’s the difference between the matching score and the AI score?
A: The matching score is the percentage of text that matches internet sources and the internal database (plagiarism/similarity). The AI score is the percentage of text that Copyleaks identifies as AI-generated content. They are calculated separately and serve different purposes.
Q: Will using writing assistance suggestions flag work as AI?
A: Simply making manual changes based on human-readable suggestions (for example, a teacher’s comments, simple style tips, basic changes like spell and grammar corrections) does not by itself cause AI flags. Copyleaks focuses on whether the text is being rewritten or generated by an AI system, not on normal manual editing.
Q: Can faculty resubmit work for students?
A: I would change this to "Faculty can re-submit by impersonating a student or uploading the content to the Copyleaks Teacher Scan Tool. The exact workflow can vary depending on your institution’s Canvas configuration.
Q: Does Copyleaks work on Safari on a Mac?
A: No. The preferred browser is Chrome.
Q: Does Copyleaks work on an iPad?
A: Yes, on the Chrome browser, you can successfully access Copyleaks tools. If you are receiving a “Bad Page Request”, Copyleaks recommends the following steps:
· Use Google Chrome (preferably in Incognito Mode).
· Clear all cookies and cache.
· Verify that cookies are enabled.
Q: How accurate is Copyleaks’ AI detection, and how is it different from comparing a student’s submission to known sources?
A: Copyleaks says its AI detection is about 99.8% accurate, based on their internal testing and third-party evaluations. AI detection and plagiarism checking are separate:
· Plagiarism score compares the text to internet sources and the shared database for identical or slightly changed matches.
· AI score uses a model that looks for patterns typical of AI-generated writing, not matches to known texts. The two scores can overlap, but they are calculated independently.
Q: Does AI detection also flag students who write in Spanish and then use a translator to convert their work into English (often plus Grammarly)?
A: Yes, that will almost always be flagged as AI. Translation tools and advanced grammar tools (like Grammarly when used beyond basic spell-check) rely on AI, so the resulting English text typically looks AI-generated to Copyleaks and will be marked as such.
Q: Will students be able to see the color-coded similarity flags and manipulation alerts, or are those for faculty only?
A: Students do see the color-coded similarity flags in the report, and they also see any alert about suspected manipulation tactics.
Q: Will there be a document that shows all of the recommended settings?
A: Copyleaks doesn’t yet have a detailed “recommended settings” guide for every assignment type, but many options in the interface are labeled as “recommended”. What are the recommended admin settings for my institution?
Q: Will most of these settings be on by default, or do we have to set them every time?
A: Most of the settings instructors care about will either be defaulted or can soon be saved as default scan settings for assignments. Once that feature is fully live, you’ll set your preferred scan options once and reuse them when creating new assignments. Typically, the only thing you might tweak assignment-to-assignment is what content you omit (e.g., references).
Q: If a student submits twice to the same assignment, will Copyleaks scan both? What appears in the report?
A: Yes, each submission gets its own report. In Canvas SpeedGrader (and in the student view), the most recent submission is shown by default, but you can still access the earlier report. Copyleaks does not compare those two submissions to each other, so that second attempt isn’t treated as plagiarizing the first.
Q: In our HUM1020 course, students upload a first draft and a final draft to two different dropboxes. Copyleaks marked all final drafts as 100% copied. Why? Can it tell the same student submitted both drafts?
A: If the drafts and final are in different assignments and the shared database is on, Copyleaks treats the first draft as a stored source. When the final draft is submitted to a different assignment, it gets compared against that stored draft and may show as 100% copied, even for the same student.
Right now Copyleaks cannot manually exclude other assignments from comparison.
Q: In SpeedGrader, can we add a rubric like we normally do, or use a rubric directly inside the Copyleaks report?
A: Inside the Copyleaks grading view, you can’t use a fully integrated rubric the way Canvas does. You can attach a rubric as a file in the comments area:
· Upload your rubric document as a comment attachment under grading/feedback.
· Students get a notification and can click the attachment to view the rubric.
Q: We still have the same issue as Turnitin where the AI score doesn’t show in SpeedGrader until you open the full report. Is that correct?
A: Yes. Currently, the preview panel only shows the plagiarism/similarity score. To see the AI score, you must open the full Copyleaks report. Canvas is planning an update (via a new asset processor) that should allow multiple scores in the preview, but that’s targeted for a future release (mentioned as around summer 2026).
Q: Can we exclude references in the assignment settings?
A: Yes. In the Copyleaks settings for an assignment, the omit/exclude options let you exclude references (and other sections you choose). Anything set to be omitted is not checked for plagiarism or AI.
Q: Does Copyleaks check whether references are real or fake?
A: No. Copyleaks does not verify whether references are real or fabricated. You would need to spot-check references yourself if you’re concerned about “hallucinated” citations, even if those references appear in the report.
Q: Some students say they were flagged just because they “write too well.” Does Copyleaks simply treat good writing as AI?
A: No. The system is not designed to flag work just because it’s polished. The detector uses dozens of signals—including metadata and pattern analysis—beyond writing quality.
False positives can occur, especially with shorter pieces, but Copyleaks aims for a very low false positive rate. If a student insists their work is original, the recommendation is to:
· Review the report carefully,
· Talk with the student about their writing process, and
· Use the report as one piece of evidence, not sole proof.
· Review the Copyleaks Best practice guide for more information
Q: Does using Grammarly to correct grammar cause an entire essay to be flagged as AI-generated?
A: It depends on how Grammarly is being used:
Basic Grammarly (simple spelling/grammar corrections) generally does not trigger an AI flag.
Premium/advanced Grammarly features that give rewrites, wording suggestions, and style changes are AI-driven and can cause the essay to be flagged as AI.
Because of this, Copyleaks recommends steering students toward the built-in Writing Assistant instead of Grammarly for grammar help.
Q: Does the same answer about AI flags apply if students use AI-based translators (e.g., writing in Spanish and translating to English)?
A: Yes. AI-based translators and advanced grammar tools are themselves AI writers, so the result often looks AI-generated and is flagged as such.
Q: Would letting students see their own Copyleaks reports right after submitting (e.g., discussion posts) make them less likely to use AI to cheat?
A: The presenter agrees and strongly recommends letting students see their reports. When students see clearly that AI or plagiarism has been detected—and understand why—many become less inclined to rely heavily on AI tools. The report should be used as a learning tool, not just a punishment.
Q: How can we check discussion-board posts for AI, since Copyleaks isn’t directly integrated into discussions yet?
A: For now you must use a workaround:
1. Copy the discussion response text.
2. Paste it into the Teacher Scan Tool in Copyleaks.
Canvas is expected to add better discussion-post integration in a future asset-processor update, but it’s not available yet.
Q: Will the AI detector ignore quotations or other text that I’ve excluded in the settings?
A: Yes. Any text you omit/exclude (e.g., quotations, references) is completely ignored for both plagiarism and AI detection, including the “reviewed phrases”/AI phrase highlights.
Q: Why is Copyleaks assumed to be superior to other AI detectors?
A: According to the presenter:
Copyleaks claims the lowest false positive rate on the market.
They’ve used AI for years in their plagiarism detection (to understand meaning, not just exact matches).
They focus on explaining why something is flagged (AI logic, AI overview, phrases), not just giving a raw score.
They acknowledge no detector is perfect, but they argue Copyleaks is “closest” in terms of reliability and transparency.
Q: Will settings be made by course designers, and if my courses are externally designed, will Copyleaks behave like Turnitin (automatic in the master course)?
A: That’s handled at your institution:
For MDC Online, instructors are advised to contact their instructional designer.
If Turnitin settings are pushed from the master shell, Copyleaks can be set up in a similar way, but this decision is made by course designers/administration, not Copyleaks.
Q: On a printed report, will Copyleaks show which AI program (e.g., ChatGPT, Claude) the student supposedly used?
A: No. Copyleaks does not identify which AI model generated the text. It can only indicate that the writing is likely AI-generated, not whether it came from ChatGPT, Claude, Gemini, etc.
Q: Is it fair or transparent to hide AI details from students—like accusing someone of a crime without letting them see the evidence?
A: The presenter agrees with the concern and recommends allowing students to see their Copyleaks reports. While Canvas allows instructors to hide the report, Copyleaks’ position is that students should see the report as part of a learning process, not just as hidden evidence in an accusation.
Q: Is there any way to get a consolidated view of all reports instead of opening each one individually?
A: Yes. Use the Analytics view in the Copyleaks/Canvas integration:
After all submissions are in, go to the assignment analytics dashboard to see all scores in one place (AI/plagiarism), rather than opening each individual report.
Q: I submitted the same essay to Turnitin (100% AI) and Copyleaks (0% AI). Why the difference?
A: Copyleaks can’t explain Turnitin’s behavior; they don’t know how Turnitin’s AI detector works. They emphasize:
Different detectors can give conflicting results.
They do not recommend using multiple AI detectors because it leads to confusion (“which do I trust?”).
They can only stand behind Copyleaks’ own testing and false-positive rates, not other products.
Q: Is Copyleaks consistent? If the same paper is submitted again, will it produce the same report?
A: If everything is identical—same text, same file, same format—then yes, it should usually be consistent. But results can change when:
· The file format or formatting changes (PDF vs. copied text, headers, bullets, etc.).
· The student makes even small edits (e.g., changing 20–30 words in a 300-word piece can dramatically affect the AI score).
· The sensitivity level is adjusted.
So minor changes, especially in shorter pieces, can produce very different AI scores.
Q: Can Copyleaks accept formats beyond PDF and Word—like Apple Pages?
A: Yes. Copyleaks supports more than just PDF and DOC/DOCX. Apple Pages files are among the supported formats. A full list of compatible file types is available from Copyleaks’ technical specs.
Q: Can Copyleaks integrate with third-party assignment tools (beyond standard Canvas assignments)?
A: The presenter believes Copyleaks is not currently compatible with most external/third-party assignment tools, but they are not certain and would have to confirm with the technical team. At present, the supported use is via standard Canvas assignments and the Teacher Scan tool.
Q: If we add Copyleaks tools (like Teacher Scan/Overview) to a shell course, will they carry over into other sections when we copy the shell?
A: The presenter isn’t sure and notes they are not a Canvas expert. They’d need to test how shell courses behave, so instructors should check with their local Canvas support or instructional designers.
Q: Does the built-in grammar report just make suggestions, or does it automatically correct students’ writing like Grammarly?
A: It only makes suggestions; it does not make automatic corrections. There’s no button to “accept all changes.” Students must:
1. Read the grammar feedback,
2. Manually revise their work, and
3. Resubmit if needed.
This is intentional so you can see how much correction was needed and ensure students are doing the actual rewriting.
Q: Is Copyleaks’ built-in Writing Assistant as robust as tools like Packback?
A: No. The Copyleaks Writing Assistant is meant for basic grammar corrections, not full writing coaching. For more robust, AI-driven writing support, the presenter suggests using tools like Packback alongside Copyleaks.
Q: What if a student uses an online tool that emails them a “mission statement” or similar text based on their answers, and then they turn that into an assignment? Could that be falsely flagged as AI?
A: Yes, that’s possible. Two factors can lead to high AI scores here:
1. The external site may be using AI to generate the mission statement text.
2. Short submissions are harder to evaluate reliably. Copyleaks recommends at least ~350 words for more accurate AI detection; shorter pieces have a higher false-positive risk.
In such cases, Copyleaks urges instructors to review the context, talk with students, and use the report as a tool—not as automatic proof of misconduct.
Q: Is there a minimum length recommended for reliable AI detection?
A: Yes. Copyleaks recommends that submissions be at least about 350 words for better AI-detection accuracy. Below that length, the model has less data to work with, and the chance of misleading results (including false positives) is higher.
Q: Does Copyleaks ever automatically correct student writing?
A: No. The tool highlights grammar issues but never applies automatic corrections. Students must revise their work themselves and then resubmit.
Q: Which browser is recommended when using Copyleaks in Canvas?
A: Google Chrome is the recommended browser. Many reported “bad page requests” or access issues are actually browser-related, and switching to Chrome typically resolves them.
Q: Is Copyleaks compatible with third party assignments such as Cengage MindTap?
A: Copyleaks is not compatible with third party assignments such as Cengage MindTap.
Q: How do I manage Drafts and Final Submissions in Copyleaks?
A: Communicate Both Deadlines in Instructions
Create the assignment with the final due date as the official deadline in the assignment settings in Canvas.
In the syllabus and assignment instructions, clearly state:
· Draft due date: [Insert date]
· Final submission due date: [Insert date]
No need to adjust assignment settings—just communicate both deadlines to students.
Then you need to create a manually graded column in the Gradebook to enter grades for the draft version.