Can Universities Detect Chat GPT?

As an AI-based language model, ChatGPT has become quite popular among people for various reasons, including academic use such as writing essays and coding. However, the question remains, can universities detect ChatGPT?

In this article, we will explore the methods used by universities to detect ChatGPT and whether using it for academic purposes is ethical or not.

Why are Professors Adopting ChatGPT?

Before we delve into the detection methods, let us understand why some professors and educators are intrigued by ChatGPT for their teaching or research. ChatGPT has several advantages that make it appealing for academic use:

• ChatGPT can generate high-quality text on a wide range of topics and domains that can help professors in crafting lecture notes, slides, tests, or exercises.
• Professors can use ChatGPT to assist and guide students facing difficulties with their studies or ventures.
• ChatGPT can simplify the process of grading, summarizing, or researching for professors by automating some tasks.
• Professors can use ChatGPT to generate syllabi or suggest readings that are relevant to a precise subject matter.
• ChatGPT can help professors translate some of the dense or complex lessons into simpler language.

Can Professors Detect ChatGPT Essays?

One of the most common academic uses of ChatGPT is writing essays. Students can easily be tempted to use it as a quick way to complete assignments without doing much effort or research.

However, this practice is very unethical and dishonest and violates the academic integrity policies of most institutions.

It also won’t be as effective as it seems, as professors can often detect ChatGPT essays using various methods.

Plagiarism detection tools such as Turnitin and Copyscape can spot if your essay is generated by ChatGPT without any editing or citation.

Stylometric analysis tools such as JStylo-Anonymouth and WritePrints can detect the linguistic differences between ChatGPT essays and your previous writings, which can also raise suspicion among professors.

And sometimes, it might not even be necessary to use plagiarism detector tools. All it takes is some human evaluation and common sense to spot irregularities or inconsistencies in essays.

For example, professors can notice sudden changes in vocabulary, tone, or complexity which could indicate that the essay was generated by ChatGPT.

They can even ask students to explain or defend their essays orally or in writing, to test their level of understanding and involvement.

Can Universities Detect Chat GPT Code?

Another possible academic use of ChatGPT is writing code. Students can be tempted to use it to generate code out of thin air or modify existing code without citing sources, perhaps to complete their projects faster and with less effort.

It may seem like a convenient way to write code quickly, but this practice is not only unethical, like essays, it can also be detected by various methods universities employ.

Code plagiarism detection tools like MOSS or JPlag can be used to compare your generated code with a large database of online sources. If the code is generated by ChatGPT, it will be flagged as plagiarized.

Even if it is partially generated and mixed with other sources or original code, similarities will still be detected.

Code quality analysis tools such as SonarQube or CodeClimate can also help detect ChatGPT code. These tools measure the readability, complexity, documentation, testing, and other aspects of the code.

If it is generated by ChatGPT, it will have low quality and maintainability scores that can indicate its artificial origin.

Even Turnitin claims that its technology is proficient enough to spot texts generated by ChatGPT. It states that the tool has been trained on academic content collected from a massive database.