The Student Room Group

will university find out in future if someone uses CHATGPT

In the future could a uni find out if you used CHATGPT for a general idea plan by asking the creator of the bot, . Or is this all privacy info and they cannot find out. I am asking this generally because I know people who use iy
Reply 1
I meant I know people who use it so in the future could the uni find out that it was used in general even for a plan? or it it private info that they cannot obtain as phone number and email is used. i was wondering if a uni could access all this info in the future. there are may people i know who use it so how is it fair?
Reply 2
Lecturers are not stupid, they can tell if the answer is scripted, by checking your previous submissions etc.
Original post by Anonymous
In the future could a uni find out if you used CHATGPT for a general idea plan by asking the creator of the bot, . Or is this all privacy info and they cannot find out. I am asking this generally because I know people who use iy

Hey there!
As a general rule of thumb, they are saying that the plagiarism checkers can detect ChatGPT so not to use it. I think the percentage they gave was something like 98% accurate for checking plagiarism from ChatGPT. I'm not sure how true that is and whether it's as accurate as they say but that's the general consensus on using it. So even for using it in the initial stages, if it detects anything similar in the final product it will pick it up.

Hope this helped!
Lucy - Digital Student Ambassador SHU
Reply 4
ChatGPT also has a very distinctive "voice" which has become very recognisable very quickly so unless stuff generated by it is majorly rewritten, it will stand out like a sore thumb.
Original post by hallamstudents
Hey there!
As a general rule of thumb, they are saying that the plagiarism checkers can detect ChatGPT so not to use it. I think the percentage they gave was something like 98% accurate for checking plagiarism from ChatGPT. I'm not sure how true that is and whether it's as accurate as they say but that's the general consensus on using it. So even for using it in the initial stages, if it detects anything similar in the final product it will pick it up.

Hope this helped!
Lucy - Digital Student Ambassador SHU


I think most universities are avoiding using these ChatGPT checkers as they simply aren't reliable enough and so can't be used to make any decisions as to whether a student cheated or not.
OP, you shouldn't need to worry about this because you won't be using ChatGPT to do your assignments, right? RIGHT?
Reply 7
Original post by Uni_student3132
I think most universities are avoiding using these ChatGPT checkers as they simply aren't reliable enough and so can't be used to make any decisions as to whether a student cheated or not.


I don't know have you heard about Turnitins detection, but this plagiarism checker becomes really powerful, so I guess every teacher gonna use this soon. And yeah, nowadays none of paraphrasers can't bypass this Turnitins detection.
Reply 8
Original post by BellaKel
I don't know have you heard about Turnitins detection, but this plagiarism checker becomes really powerful, so I guess every teacher gonna use this soon. And yeah, nowadays none of paraphrasers can't bypass this Turnitins detection.


sorry, I dont think so.. I have already tested Turnitins detection by using Netus ai paraphrasing tool. Idk how, but Turnitin isn't able to recognize paraphrased ChatGPTs generated content when it went through Netus paraphraser. U can try this by yourself :biggrin: check https://netus.ai/
Universities can tell if it has been usesd due to its distinctive writing style, without having to get a plan of the essays. Your 'friends' will undoubtedly be found out
Original post by BellaKel
I don't know have you heard about Turnitins detection, but this plagiarism checker becomes really powerful, so I guess every teacher gonna use this soon. And yeah, nowadays none of paraphrasers can't bypass this Turnitins detection.

Most UK universities have said they won't be using it, at least for now as they can't guarantee it is accurate enough. There are also concerns about bias against students whose first language isn't English.
(edited 11 months ago)
Original post by Anonymous
In the future could a uni find out if you used CHATGPT for a general idea plan by asking the creator of the bot, . Or is this all privacy info and they cannot find out. I am asking this generally because I know people who use iy


The university know. Lecturers use AI checkers available in the Turnitin. Avoid using Chat Gpt if you can't paraphrase and also it gives 75% irrelevant arguments. Find an academic writer help you with your work. I am readily available to help. You can also work hard and up your studies. Avoid AI for some courses
Use Chatgpt and then use Grammarly. Did that for some of my final work and got away with it as Grammarly makes it sound human and gets rid of the repeated words and phrases that give it away to AI checkers. Also pay for Chatgpt 4 as it is defo worth it cause it rarely lies anymore.
Reply 13
Original post by Frank2003
sorry, I dont think so.. I have already tested Turnitins detection by using Netus ai paraphrasing tool. Idk how, but Turnitin isn't able to recognize paraphrased ChatGPTs generated content when it went through Netus paraphraser. U can try this by yourself :biggrin: check https://netus.ai/


That's right, turnitin gets confused when you feed it through multiple generators at the moment. But guess what? We markers aren't as dumb as turnitin's algorithm - it is obvious to us when you've used AI to write your essay, and we don't take that kindly to it. It's kind of demoralising marking something that the student hasn't even tried to write themselves. So you are unlikely to get a very warm response. However - top tip - if you write your essay yourself we love that. Warts and all. Much better than generic algorithm-generated nonsense. It's actually easier than ever to get a first now - if you write your essay yourself rather than using AI, it will stand out from the bunch and the marker will be so relieved to read something obviously written by a human that they'll probably give you a first regardless of how bad your argument is!
Reply 14
Original post by Anonymous
Use Chatgpt and then use Grammarly. Did that for some of my final work and got away with it as Grammarly makes it sound human and gets rid of the repeated words and phrases that give it away to AI checkers. Also pay for Chatgpt 4 as it is defo worth it cause it rarely lies anymore.


You know that Turnitin's AI-detection works retrospectively, right? So as they update and improve it, it will scan old submissions and catch stuff that slipped through the net in earlier iterations?
Reply 15
Original post by Uni_student3132
I think most universities are avoiding using these ChatGPT checkers as they simply aren't reliable enough and so can't be used to make any decisions as to whether a student cheated or not.


Some universities (I know of one) have opted out, not because it isn't reliable, but because they'd rather not know. The AI-detection on Turnitin is already highly reliable in its positive scores (the false positive rate with a high score is close to nil on the evidence so far), but produces a lot of false negatives, which is why markers will not be taking a low score as evidence that an assessment isn't written by AI. They will use their judgment, and it is very obvious to an experienced marker when something has been written by, or with the aid of, AI.

Quick Reply

Latest

Trending

Trending