Identifying AI use in misconduct cases

Providing evidence for unauthorised use of Artificial Intelligence (AI) may be difficult when lodging academic misconduct cases. Learn more about how to use the Turnitin AI indicator with caution, and other indications of AI use.

Under the ACU Student Academic Integrity and Misconduct Policy, unauthorised use of generative artificial intelligence (genAI), paraphrasing and translation tools are not permitted.

Suspected instances of unauthorised use should be lodged via SAMMS by the lecturer-in-charge, with appropriate supporting evidence. However, identifying appropriate evidence may be difficult in suspected AI cases.

How should I use Turnitin’s AI writing detection tool?

Sometimes the first indication that AI has been used is the Turnitin AI indicator tool. It provides an overall percentage that represents the proportion of the work that the tool is confident was written by genAI (with 98 per cent confidence). It also highlights the text that it believes was AI generated; this is separate from the text matching for the similarity report.

This tool may be useful as an initial ‘flag’ to indicate that genAI may have been used, but due to the risks of false positives, it should not be used as the sole evidence to forward an allegation of potential academic misconduct.

Cases being referred for investigation must contain additional evidence, especially for cases where the AI flag is lower than 20 per cent, as low percentages are associated with a higher risk of false positives. False negatives also occur and a low (or zero) AI indicator percentage does not necessarily mean that there is no AI generated text.

Additionally, work that is not submitted through Turnitin can also be referred for review, without an AI report, if there is evidence to support suspected AI use.

What other indications of AI use should I look for?

• fabricated references
• fabricated quotes
• incorrect information
• material which is slightly off topic, repetitive or has the wrong focus
• very short editing times (with no drafts)
• absence of typographical errors

Additional guides around suspicious features to look for, and how to provide high quality evidence are:

• Turnitin AI indicator checklist
• Best practice supporting evidence

Want to learn more?

The Centre for Education and Innovation (CEI) will be hosting two online workshops on 'Artificial Intelligence: methods to detect inappropriate use'. Register now to receive a calendar invite for one of the workshops below:

• Tuesday, 28 May, 10am-11am
• Wednesday, 29 May, 2pm-3pm