Cheating in exams with artificial intelligence
Complaints of teachers in Lebanon are increasing that students rely on GBT chat to cheat in exams (Pixels)
A sudden student brilliance baffles teachers and university academics
As if school students were lacking a modern innovative way to facilitate cheating in tests, assignments, and exams before human artificial intelligence applications were carried to them on a plate of technology to tickle the imagination and tricks of some students, and to confuse professors and teachers in schools and universities with the matter of “savvy” and the organization of ideas and the integrity of language suddenly falling on Students entrusted them during the academic year with a different performance.
Old generations remember cheating in exams with deviant eyes and hesitant movements, and when daring increases, anticipation and waiting rates rise before the student opens (a pinch) any piece of paper written on it some of what he did not memorize, or an equation, or pen tips that guide him to the correct answer.
The methods of cheating have always changed from writing on rulers, erasers, tables and under chairs, to rolled scraps of paper hidden in the corners of clothes or shoes, to writing inside shirts and school uniforms, on the insides of belts or on parts of the body, or inside a pencil case or calculator cover. And other places that every reader of this article can add according to the available means of his time.
This is of course except for the cheating that comes without preparation, and is in the form of peeping at the answers of other students or asking them to answer in a whisper or in sign language if the monitoring conditions allow.
As soon as the mobile entered the examination classes, many cases of cheating were detected first through text messages, then with smart copies of it in the texts of e-mail messages (e-mail), then using printed or photocopied documents saved to Bluetooth headsets, smart watches and the like.
Faced with the arrest of many cases after manual cheating turned into electronic, the ministries of education in several countries established many controls in examination halls, preventing the introduction of phones, surveillance cameras in the classrooms, and students searched with devices such as the electronic stick in Egypt, China, Morocco and elsewhere.
sudden intelligence
The Lebanese teacher, Samar, tells that she was lucky because her students were under eight years old, and the technology of artificial intelligence had not yet reached them, after she heard her colleagues complain about the task of making sure that their students actually solve their tests themselves, as after parents intervened to solve their children’s homework, the students became dependent on applications and websites of artificial intelligence.
Samar adds that she tried some of these sites and used them to find new ideas and questions for students in comprehension.
In a different context, Basma Harb, a middle school teacher, says that she suddenly noticed that her students had advanced in the level of writing in some assignments, only to discover later that they had taught each other to use "Chat GPT" to write composition texts, and the French language was missed in their homework correctly. One hundred percent, which prompted her to suspect a group of students, only to discover after investigating them and communicating with their families that they had cheated. It was only for her to repeat the assignment in a school hour after stripping the students of their phones.
Basma says that she had not tried any application of artificial intelligence before, but she found it useful in preparing for exams and lessons, and it may be used as an assistant for the teacher to shorten research time as well.
Basma drew the attention of her colleagues to the matter, so that the mathematics teacher discovered that most students solve problems using artificial intelligence applications, especially in algebra, arithmetic, and calculus, as they noticed the matter when they found the students finishing the competitions at a record speed.
university procedures
The academic at a Lebanese university, Maryam Kurdi, says that she revealed some students cheating after submitting a mid-year project through a program that shows whether the writing is human or automated, to find that more than 60 percent of the students used artificial intelligence applications.
She adds that some of them circumvented and changed some sentences and the method of numbering and indexing, but in the end they were revealed in different ways, explaining that she is not against using these applications but in a scientific way, and not copying texts without understanding them and verifying information.
At the same time, Kurdi prevented the introduction of any electronic means in the exams at the end of the year, and also placed all cellular devices and smart and non-smart watches belonging to the students on the table in front of her during the final exams.
Applications disclaim
In a survey conducted by Best College last March on a thousand university students, it found that 43 percent of students have experience using artificial intelligence applications, and that at least one in five students used them to complete their homework, while nine out of 10 Students experiment with his tools on personal projects out of curiosity.
This survey showed that 57 percent of university students do not intend to use it or continue to rely on it to complete the tests, yet a third of them do so, and 11 percent prefer not to answer, although 51 percent of the students surveyed considered using artificial intelligence tools. in their jobs constitutes fraud or plagiarism.
For its part, the company "Open AI", which designed "Chat GBT", considered that its program should not be used for misleading, which made it develop the program in a way that allows it to know whether the texts were generated by the program itself by machine or human intelligence, in order to prevent fraud in all cases. fields, and the use of this technology in an honest manner.
How does it detect artificial fraud?
Students use AI to cheat on exams and projects by either creating complete texts, or answering questions. This can be useful for students stuck in a particular problem, and a helper for research, such as using electronic search on "Google" or similar sites without falling into plagiarism, but using it for fraud and transfer gives student users an unfair advantage or mark.
Professors can use AI-powered tools to detect plagiarism to prevent students from cheating by copying content from other sources, but at the same time students can also use AI-powered plagiarism detection tools to identify content they can copy without being caught.
From here, teachers must inculcate a culture of academic integrity to discourage students from cheating, and make them aware of the importance of academic honesty that develops science and increases knowledge in authentic dimensions that are not fabricated or copied.
Teachers must also be aware of the latest artificial intelligence cheating techniques to confront this scourge, so they subject research projects and assignments to investigation through programs that detect machine writing for human writing, especially if the writing patterns are very perfect.
Perhaps the most prominent of these applications (Originality.ai, GPTZero, Copyleaks) and others that detect plagiarism and work on natural language processing and machine learning to detect whether the content was generated by artificial intelligence or if it is the product of humans.
Source : websites