How to detect AI Generated Essays
Darren Coxon
Founder @ Node | AI Strategy, Training, Policy and Tools for the Future of Education
In my first article on AI in education , I outlined concerns many have over the increase in AI generated content. The release of Chat GPT in November, with its ability to create high quality, human-like written text on any subject you care to test it on, has many of us in education concerned.
However, there are already ways you can detect whether or not your students are using AI generated content. One of the easiest is using an online AI content checker. There are a few of these around, such as the one from AI marketing company Content at Scale: Free AI Detector - AI Content Detection for GPT-3 + ChatGPT (contentatscale.ai)
You simply paste in the text and it will tell you the likelihood of it having been generated by AI. For example, an essay Chat GPT has created on Macbeth's witches brought a 1% human content score:
However, if I drop in some exemplar text from an AQA mark scheme, it shows me the opposite:
领英推荐
I would strongly suggest speaking to your students at the start of term, explaining that any essays created using AI will be sanctioned in the same way as plagiarised essays. I would also suggest rewriting your academic honesty policy to include AI generated content.
If you run an essay through this detector and it's either clearly AI generated or perhaps borderline, you will still need to triangulate by asking the student questions on their essay. I would advise against immediately sanctioning, just in case the detector has got it wrong. It's highly unlikely, but certainly worth checking.
Of course, if your student has been getting grade 4 or 5s, then suddenly pulls a grade 7 or 8 essay out of the bag, this is likely going to make any teacher suspicious. But having the above tool at your disposal does mean that you can use the same sort of technology to your advantage, and hopefully stop AI generated content from plaguing your classes.
As I mentioned in my first article on the subject, it is not all bad news, and in future articles I will explore how we can use these new AI technologies productively for both our own benefit, and that of our students. After all, it's not going anywhere, so we have a duty to educate, not to ignore.
Monitoring And Evaluation Coordinator at U.S. Embassy Pristina
1 年Yes, they will be much needed. One case that i know is Crossplag?, you can check it out for free here: https://crossplag.com/