I was falsely accused of using generative AI for an assessment task. Despite providing proof, including evidence of a draft that I used to generate my ideas, I was referred to the academic misconduct department. This team has a backlog of a year’s worth of cases, causing delays in students' course completion and academic transcripts. Shockingly, I was flagged again in a second assessment and made sure to provide evidence that I did in fact generate my ideas and my assessment myself. Proving that I did not use AI has become extremely difficult as there is hypocrisy in relying on AI (which education providers consider inaccurate themselves when it comes to student use) to detect generative AI. The burden of proof now relies on the student that the detection tool is in fact flawed.

I’ve also tested the reliability of TurnItIn’s AI detection tool on essays handed in before the age of generative AI - and sure enough, those old essays from 2017 were flagged for up to 70% AI assistance.

The huge backlog of cases in the academic misconduct team, especially with the introduction of generative AI, demonstrates an issue of over-reliance on tools which should not be solely relied on to detect academic misconduct. TurnItIn’s website does state that the AI detection tool is not 100% accurate and is still in the process of being developed.

Many hardworking students like myself have been falsely accused of using AI in assessments. It is a global issue and will only continue to get worse until education providers establish ways to detect academic misconduct that is accurate and truthful. The inaccuracy of this tool also raises the issue of students also actually using generative AI without being detected. Alternatively if this is too challenging, education providers should recognise that with the increase of AI in education, workforces and technology - there needs to be a way to utilise AI in education whilst developing traditional critical thinking and problem solving skills.  

TurnItIn's AI detection tool is flawed and can misidentify legitimate student work as generated by AI. This not only affects students' academic records but also their mental health and future opportunities.

Sign this petition to urge education providers such as Western Sydney University, University of Sydney, UNSW and UTS to stop using TurnItIn’s unreliable AI detection tool until it is proven 100% accurate and fair!