The problem of AI-based cheating is currently a hot topic, pushing many universities to review their assessment policies.
A well-known essay plagiarism detector is allegedly developing a brand-new tool to find if students have used ChatGPT or other AI bots— and it is already being used.
The fluidity of ChatGPT, which can produce essays, poems, prompts, contracts, lecture notes and computer code, has astounded people, though occasionally at the expense of accuracy or creativity.
Thousands of students from all around the world have shared their experiences using the AI system in their studies, with many opting to let the impressive technology write a whole essay with accurate and persuasive information including citations, with just a click of a button.
The ability of such bots to shake the education system has pushed many professors and academics to question the way education is now being provided. Though AI has served as a revolutionary tool in many aspects, many feared the outcome of it on academia.
When ChatGPT was released months ago, even the most comprehensive plagiarism detectors could not detect the AI’s work when it comes to essay writing.
However, now as the technology continues to gain more popularity among students, programmes are starting to create their own AI Innovation Lab to outsmart the bots and detect the use of AI writing tools.
Last month, Turnitin developer David Adamson provided a demonstration of the new software’s functionality in a YouTube video. After inputting an essay that ChatGPT had written and edited himself, the system determined that 50% of it was created by AI, even identifying 22 of the essay’s 43 sentences as not being human-written.
“Our model has been trained specifically on academic writing sourced from a comprehensive database, as opposed to solely publicly available content,”Turnitin’s chief product officer, Annie Chechitelli, explained in a blog post.
“As a result, Turnitin is more tuned to finding instances of potential dishonesty in student assignments.”
Although the company’s current products already provide some level of AI detection, CEO Chris Caren continued, more would be required in the future.
“We will incorporate our latest AI writing detection capabilities — including those that recognize ChatGPT writing — into our in-market products for educator use in 2023,” he said.
The team is hoping that the tool, which is currently in testing, will be ready soon. Some students have already revealed in Tiktok videos that their essays were detected for plagiarism after using AI technology, hinting that the trails have proven to be successful.
Despite the improvements however, some have come up with even better AI systems that would make it harder for TurnItIn to outsmart the bot.
Now, it is a matter of determining who will be faster with technology developments.