Versions Compared

Old Version 5

changes.mady.by.user Franklin Clark

Saved on

compared with

New Version Current

changes.mady.by.user Franklin Clark

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Info

SummarySome students might use ChatGPT or other artificial intelligence (AI) tools to generate writing assignments. Beginning April 4, 2023 Turnitin will modify its similarity report to highlight text that may have been generated by AI

Turnitin's AI writing detection capability is designed to help educators identify text that might be prepared by a generative AI tool. The AI writing detection model may not always be accurate (it may misidentify both human and AI-generated text), so it should not be used as the sole basis for adverse actions against a student. It takes further scrutiny and human judgment in conjunction with an organization's application of its specific academic policies to determine whether any academic misconduct has occurred.

Instructions

  1. To use Turnitin’s AI detection feature, simply follow your usual similarity checking workflow.

  2. Once the submission is processed, click on the similarity report.

    1. The AI score is not embedded directly in the Canvas user interface, so you will need to view the similarity report to see the AI score and to access the AI report.

  3. The AI indicator is linked to the AI report.

    1. The AI report highlights the text segments that the Turnitin model predicts were generated by AI.

    2. The AI report shows a percentage of text predicted to have been generated by AI.

  4. See this Turnitin article for details on interpreting the AI writing indicator and report.

Considerations

  1. The AI indicator and report is visible only to instructors and admins.

  2. Turnitin has created an expanded FAQ related to AI detection; if your question isn’t answered in this document, please submit a help desk ticket.

  3. Submissions made before April 4, 2023 will need to be re-submitted if you wish to check for possible AI-generated text in those submissions.

  4. Turnitin’s current detector is trained for long-form English content and can detect GPT3/3.5 and ChatGPT signatures.

    1. The model does not reliably detect AI-generated text in the form of non-prose, such as poetry, scripts, or code, nor does it detect short-form/unconventional writing such as bullet points, tables, or annotated bibliographies.

  5. File requirements:

    1. File size must be less than 100 MB

    2. File must have at least 150 words of prose text in a long-form writing format

    3. File must not exceed 15,000 words

    4. File must be written in English

    5. Accepted file types: .docx, .pdf, .txt, .rtf

  6. The AI detector’s false positive rate (incorrectly flagging human-written text as AI-generated) is claimed to be less than 1%, but it isn't zero percent. To help navigate conversations around false positives, in addition to the response provided in the expanded FAQ document liked above, please use these resources:

    1. Understanding false positives within our AI writing detection capabilities

    2. How to prepare for and discuss the possibility of false positives

    3. AI conversations: Handling false positives for students

  7. Turnitin has created an expanded FAQ related to AI detection; if your question isn’t answered in this document, please submit a help desk ticket.

Page Tree
rootTurnitin (Instructor)
startDepth1

...