In the modern world, technological advancements, particularly Artificial Intelligence (AI), have significantly revolutionized all aspects of human life. No one can deny the benefits of AI to the development of science, technology and countless other aspects relating to the betterment of humanity. However, the introduction of ChatGPT by OpenAI raises ethical concerns about its potential application in exam cheating.
The language model’s ability to generate human-like responses and provide instant information may tempt students to rely on it for unfair advantages.
ChatGPT is a language model for end-user interactions which operates similarly to how a search engine displays web pages. Its free version, on the other hand, excludes search features and generates responses based on training data, which leaves room for error.
In the ongoing debate over the use of AI technology in education, various K-12 school districts, notably New York City Public Schools (NYCPS) and Los Angeles Unified School District (LAUSD), have taken measures to exclude OpenAI’s ChatGPT from their classrooms. These districts have taken a firm stance against the AI approach, citing concerns about content safety and the deterrence of critical-thinking abilities.
Sciences Po in Paris and RV University in Bengaluru have joined the list of universities that have openly prohibited the use of ChatGPT on their campuses.
However, schools and universities in the United States have failed to reach a consensus. Each individual institution is left with the discretion of deciding whether to ban this innovation or not. A recent poll found that only one-third of college students reported AI tool bans at their educational institutions.
As the use of artificial intelligence (AI) grows, so do concerns about its impact on education, particularly with respect to the recognition of chat-generating pre-trained transformer (GPT) language models that students may use for cheating on tests in class or online.
Concerns about students utilizing ChatGPT to cheat have necessitated the development of ChatGPT text detectors in the form of a free tool to tackle the issue. However the research firm OpenAI issued a less-than-glowing review of this mechanism to combat such illicit behavior.
The “classifier” program adequately identifies only 26% of AI-written texts tagged as “Probably AI-written,” and it has a 9% possibility of producing false positives, mistaking human-written work for AI-created work.
A recent article by USA Today, placed the spotlight on the AI tools detector false detection weaknesses. The author says: “In another publicized case, a Washington Post technology columnist found that Turnitin’s new AI detection tool falsely found several papers written by California high schoolers to be fabricated.”
Another article published by the Washington Post revealed the same limitation about the AI detection tools. The piece highlighted the story of a senior high school learner named Lucy Goetz who received the highest grade on an original essay she authored about socialism. Surprisingly, a new instructional program erroneously pointed to artificial intelligence assistance on her paper.
ChatGPT, the artificial intelligence (AI) virtual assistant, contains inherent limitations just like any other invention, such as the inability to answer specific written queries, low quality replies, and accidental question-answering. The chatbot’s data is restricted until 2021, and the source of the answer is not provided.
It describes experiences rather than feelings, and the terminology it employs is courteous but lacks the nuances of human language, such as sarcasm and metaphors.
At the time of writing, seven countries have strictly prohibited ChatGPT. The countries are listed as follows: Russia, China, North Korea, Cuba, Iran, Syria, and Italy. Some of these nations prevented the program due to privacy concerns, while others–most notably North Korea, China, and Russia–feared that the US would use ChatGPT for spreading misinformation.
If the countries above made a stand concerning the use of this language model, Morocco; however; has not moved a muscle on this matter.
Recently, Moroccan baccalaureate students sat for the national exams, and surprisingly during the correction, teachers detected some answers surely written by the language model ChatGPT.
An illustration of this would be:
1. Are these statements true or false? Justify (3pts.)
Q: The Moroccan football team was expected to reach the semifinals?
A: False. “There is no mention in the text about the expectations for the Moroccan football team to reach the semifinals.”
Q: Many famous people congratulated Morocco for its success in the World Cup?
A: False. “There is no mention in the text about famous people congratulating Morocco for its success in the World Cup.”
While teachers in Morocco are adopting a Competency Based Education (CBE)–which is a learning approach that focuses on the results of language acquisition–it seeks to improve students’ abilities, knowledge, attitudes, and behaviors in order to prepare the learners for effective real-world tasks and applications of what they have learned.
This approach is student-centered, thereby enabling them to actively participate in the learning process and to assume responsibility for it, while also being introduced to problem-solving via critical thinking. ChatGPT impedes the development of these skills, transforming the independent behaviors into dependency and stifling critical thinking and problem-solving techniques.
When asked about her insights on the use of ChatGPT by students in cheating, Lamyae Eddai, an English teacher from Casablanca said: “As a high school teacher, I was responsible for correcting the English language national exam papers, and I can say that we noticed several cases of cheating and some of them were believed to be with ChatGPT, especially in answering text comprehension questions and writing tasks.”
When questioned on how they recognized AI answers, Eddai answered, “We knew it was AI generated due to the lack of human expressions as it is so different from Robot expressions.” She continued, “They have written something that was not mentioned in the school syllabus/curriculum and sometimes not even mentioned in the text.”
As an answer for the perils of ChatGPT use on students’ creativity, the educator stated, “AI integration in educational activities may hinder creativity and autonomy by motivating pupils to rely on algorithmic answers, limiting their ability to think critically and independently,” while adding, “It leads to ethical challenges by possibly promoting cheating and ignoring the holistic, competency-based approach to education that fosters students’ overall growth.”
Although ChatGPT models can be helpful with respect to particular tasks, their use in academic settings should be supervised and regulated to ensure that students are not engaging in academic dishonesty or cheating.
Students, in particular, need to gain knowledge of artificial intelligence in order to be competitive and effective in job markets. However, instructors, educators, and caregivers’ roles are to teach these individuals the wise and effective means of using such tools.
AI has become a reality. It is not going anywhere, and these are serious ethical issues which must be considered and discussed if we are to properly prepare students and educators for this novel phenomenon.