The "internal fate of the fire person" trend that has been spreading on Vietnamese social networks recently has attracted many mixed comments. In addition to the humorously misspelled exam title ("internal fate" instead of the correct "indignation"), the attitude of students, teachers and the way FPT Polytechnic College handled the situation, the incident also raises the issue of students abusing AI applications to cheat in exams.
AI cheating is becoming an increasingly common problem in education, as artificial intelligence tools are misused. The scope of AI cheating is not limited to any one level or field, but is widespread across both high school and college levels.
Sharing with Wired, Ms. Christina Wyman, a teacher in the US, said that AI is making cheating in the classroom easier than ever.
Automated essay generators, also known as “essay mills,” can generate complete, coherent essays based on keywords or a simple outline, allowing students to submit assignments without having to write them. Examples include EssaySoft, Articoolo, and Dr Assignment Auto Writer.
For math, apps like Wolfram Alpha, Microsoft Math Solver, and Photomath can solve problems ranging from algebra to complex physics, providing quick solutions without students having to think for themselves.
Additionally, in online exams, some AI tools can assist students by solving questions in real time or interfering with screen sharing, making it difficult for proctors to detect cheating. These tools can even adjust webcams or screen sharing to hide dishonest behavior.
Some AI tools can even simulate student behavior, such as typing or mouse movements, making it harder to detect cheating in online exams.
Professor Steve Fuller, a sociologist at the University of Warwick in the UK, often notices unusual patterns in student writing, especially when phrases are repeated in a robotic way. This can be a sign of using AI tools like ChatGPT, where answers may be correct but are too general and not directly related to the course material. According to him, cheating papers often lack references to the required course material.
Other professors have also expressed concerns about the impact of AI in education. Des Fitzgerald of University College Cork (UK) said that the use of AI by students has become "widespread" and described it as a "machine that produces poor quality content".
Another big problem for teachers is that AI detection software is not yet reliable enough, and many universities do not use it. This makes professors nostalgic for the days of traditional plagiarism, when detection software could compare text to the original document. With ChatGPT , there is no clear source document to check, making it harder to prove cheating.
One study found that cases of academic misconduct have doubled or even tripled since the advent of AI. But professors are often reluctant to report them without hard evidence. Fuller said he often had doubts but reflected them in his grades and comments.