Can Turnitin detect ChatGPT?

Can Turnitin detect ChatGPT?
Jim Casson Updated on by

Video Gamer is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Prices subject to change. Learn more

With the surge in the use of AI in everyday life, and the resulting surge in students using the tools to write part or even all of their essays, Turnitin has now announced an AI writing detection tool.

The plagiarism detection company announced they would be integrating the system, which they claim has a 98% accuracy rating, for detecting text written by an AI on April 4th.

EXCLUSIVE DEAL 10,000 free bonus credits

Jasper AI

On-brand AI content wherever you create. 100,000+ customers creating real content with Jasper. One AI tool, all the best models.
TRY FOR FREE

Copy AI

Experience the full power of an AI content generator that delivers premium results in seconds. 8 million users enjoy writing blogs 10x faster, effortlessly creating higher converting social media posts or writing more engaging emails. Sign up for a free trial.
ONLY $0.01 PER 100 WORDS

Originality AI detector

Originality.AI Is The Most Accurate AI Detection.Across a testing data set of 1200 data samples it achieved an accuracy of 96% while its closest competitor achieved only 35%. Useful Chrome extension. Detects across emails, Google Docs, and websites.

How does Turnitin detect AI writing?

Turnitin saw the issue of AI written text on the horizon and has been developing a tool to detect it for over two years now. The fruits of that labour have now been unveiled, to the potential relief of educators and dismay of students.

The company has not divulged exactly how the system works, but it is likely a pattern recognition AI of its own.

An obvious concern here is that Turnitin may create false positives, incorrectly accusing students of having used AI-generated text. A chief product office at Turnitin explained how the company has sought to avoid this saying that “to maintain a less than one per cent false positive rate, we only flag something when we are 98 per cent sure it is written by AI based on data that was collected and verified in our controlled lab environment.” 

However, educational institutions will have the option to opt out of using the tool and it looks like several Univitiers will.

The software is also, arguably, only kicking the can further down the road, as AI writing tools will only get more powerful, being able to create more natural and complex language.

Can any AI chatbots can’t be detected by Turnitin?

Turnitin’s AI detector isn’t trained at specifically detecting ChatGPT – so it’s likely there’s not many AI language models out there unlikely to be flagged by the tool.

There are, and we hope that there aren’t any lecturers out there reading this, ways to make ChatGPT undetectable.

Personally – we’ve found that AI content is incredibly easy to detect when it’s AI generated from the ground up. However – when you introduce human-written original content to ChatGPT and others such as Google Bard – it actually becomes slightly more believable. This does put your academic integrity at a risk, but at least your own original content makes up the base of the paper.

What are schools and universities doing about ChatGPT?

Back in 2022, Turnitin CEO Chris Caren stated that “educators, administrators, students, and technology partners have […] a responsibility to participate […] in the AI-assisted writing dialogue.” 

Caren also urged those within the education system to discuss and set the boundaries of acceptable use of AI in learning.

This is especially pressing right now because the new GPT-4 model can accept visual prompts.

This means it is incredibly useful for students with homework or assignments to complete, but there are also ethical implications to consider.

Hypothetically, a student could simply send an image of a diagram or question from their homework, and the AI tool could solve it in minutes.

So in an effort to stop students from submitting work that’s not their own, schools across the US and the UK have started to block ChatGPT from school networks and devices and many are likely to make use of this new Turnitin tool to further dissuade students.

How does Turnitin work?

Turnitin is a plagiarism detection software widely used in schools, colleges and universities.

The software, based on AI itself, works by scanning phrases of text from assignment submissions. Turnitin then compares them to huge archives of data. The software then tries to find matches to determine whether any plagiarism has occurred.

Text generated by ChatGPT does not exist elsewhere – the model uses the massive amounts of data it is fed in training to create original sentences and phrases. This means that Turnitin’s traditional cross-checking method would not be able to detect AI-assisted work.

The gap has seemingly now been closed, however, with Turnitin unveiling its AI writing detection tool.

Frequently Asked Questions

What does Turnitin detect?

Turnitin uses its large database of pre-existing texts to detect similarities between texts.