Artificial intelligence worries teachers, who fear that one day they may be replaced by a robot.
What are the cons and disadvantages of AI? What are the limits to the use of AI in teaching? Is this technology honest and effective?
Let's find out the answers to all these questions together 👇
Summary :
Some arguments suggest that the use of AI in teaching may lead to a lowering of academic standards. Indeed, certain behaviors and consequences raise questions.
AI can encourage an overly standardized approach to teaching and assessment, where students’ skills and performance are reduced to quantitative measures. This can overlook the diversity of talents, learning styles, and individual interests, leading to a loss of richness and creativity in education. It may foster a learning environment focused on finding correct answers rather than deeply understanding concepts, making it a real disadvantage of using AI in education.
“The recent example of ChatGPT shows us that some institutions have banned it for fear that this tool will facilitate cheating and lead to a drop in academic standards, while others have welcomed it into their classrooms as it seemed impossible to fight against these technologies - and fight a losing war. Both positions are defensible.” Should AIs be banned from schools? - Numerama
Generative AI can facilitate research and the writing of academic papers. However, when students rely solely on answers generated by chatbots or virtual assistants, they miss the opportunity to question, analyze, and develop their own ideas independently.
AI tools can encourage a tendency to focus on finding answers rather than deeply understanding concepts.
Moreover, the use of generative AI in writing can lead to uniformity in results and a loss of diversity in artistic, literary, or other forms of production. This can be damaging and constitute a serious disadvantage for AI in education.
Not all students have equal access to the technologies needed to benefit from AI, which can further deepen the digital divide, more than it is now. Inequalities between students' access to technology can be very significant, depending on whether the school is located in the city, the suburbs or the countryside. This is even more pronounced between developed countries, developing or emerging countries and underdeveloped countries.
It is crucial to ensure that AI does not exacerbate existing inequalities by guaranteeing equitable access to tools and taking measures to bridge technological gaps.
The use of AI involves the collection and analysis of personal data about students: names, email addresses, browsing history, learning preferences, academic results, etc.
This raises concerns about privacy, security, and the handling of this data. It is essential to implement data protection measures to ensure that learners' personal information is not compromised.
Users must be informed about the data being collected, how it is used, and have control over their personal information. Privacy policies and informed consent are key elements in ensuring responsible and transparent data collection.
Despite all the talk and focus on the subject, the answers from generative AI developers remain unclear. Data confidentiality remains a major challenge for the use of AI in education.
AI can collect personal data when processing natural language, such as during communication with chatbots or virtual assistants. Message exchanges or conversations may be analyzed to understand your needs, respond to your questions, or provide relevant information.
AI can also collect data from sensors and smart devices such as cameras, microphones, or wearables. These devices can gather contextual information about your environment, movements, habits, heart rate, and more. Some of this data may be sensitive and therefore pose a threat, such as medical data.
AI can also collect personal data from external sources, such as social media, public databases, and identification information provided during service registrations. This data can be used to enrich user profiles and enhance the personalization of experiences
However, this can raise concerns regarding the cons of AI in education. The collection of data from external sources can lead to issues of privacy and security, particularly when it comes to sensitive student or educational staff information.
Human learning is a complex process involving cognitive, emotional and social aspects. Although AI can be useful for certain specific tasks, it cannot yet fully reproduce the complexity of human learning. It is important to recognize the limits of AI and not to completely substitute human interaction in teaching.
Human learning is a holistic process that encompasses cognitive, emotional, social, and physical aspects. AI primarily focuses on cognitive aspects and may not fully account for the emotional and social dimensions of learning.
Human interactions, such as classroom discussions, debates, and group activities, are essential for fostering a deep understanding and practical application of knowledge.
Human learning is often reinforced by practical experience. Learners can interact with the real world, conduct experiments, solve concrete problems and apply their knowledge in real-life contexts. Although AI can simulate learning situations, it cannot fully reproduce the richness and diversity of real-world experience. AI has its limits, and cannot substitute for experiential learning.
This is indeed a limit of AI in education. Human learning is also shaped by social interactions and collaboration with peers. Discussions, idea exchanges, debates, and cooperation are crucial elements in knowledge construction. While AI can facilitate collaboration, it is essential to promote social interactions among learners to ensure a more comprehensive and diverse learning experience.
While ChatGPT and other generative AI can generate coherent responses in many cases, it can also be prone to errors and inconsistent answers. It may sometimes produce responses that appear correct but are actually false or misleading, as it lacks the ability to actively verify information.
It's important to note that ChatGPT’s training is based on GPT-4 (the latest iteration as of 2025), which has significantly improved over previous versions. However, its knowledge is still limited to data collected up until its last update, which includes information until 2023. As a result, it is possible that some information may be outdated or inaccurate, especially when it comes to recent events, research, or technological advancements. In an educational context, this can lead to misinformation.
ChatGPT is trained on a wide range of human dialogues, including conversations collected from forums, chats, and other online platforms. While these conversations help improve the model's responses, ChatGPT does not actively verify the truthfulness of the information gathered during these interactions. This can lead to biases or inaccuracies being embedded in the AI's responses, as it replicates the patterns and language seen in the training data. This remains a disadvantage for AIs.
ChatGPT draws information from various online sources, including forums, social media, and participatory encyclopedias (e.g., Wikipedia). The reliability of these sources is often questioned, as they may contain misinformation or biased viewpoints.
As a result, it is essential to approach ChatGPT's responses with caution and verify them against trusted sources. When using ChatGPT or any other information source, it’s recommended to cross-check data, consult reputable sources, and fact-check to ensure a more accurate and reliable understanding of a given topic.
The use of AI tools in education raises regulatory concerns, particularly around the authorization of students to use them. Institutions often set varying policies on AI use, with some allowing it and others restricting it. When permitted, students should cite AI usage in their work.
AI detectors can also be used to ensure academic integrity, verifying that students' work is original. As AI becomes more integrated, clear guidelines are essential to balance its benefits with maintaining academic standards.
In this article, we have explored the top 5 cons and disadvantages of AI in education. It is clear that there are many limits and challenges to address in the coming years. AI cannot replace humans.
To successfully integrate AI into education, it is essential to find a balance between AI’s capabilities and the unique aspects of human learning. Human interactions, critical thinking, hands-on experience, and collaboration must remain at the core of the learning process.
However, when used correctly, AI can offer significant benefits for education.
To find out more:
This informative article, which does not require personal reflection, was partially written with the assistance of ChatGPT. The automatically generated content has been revised (including corrections for repetition, sentence structure, added details, added citations, and fact-checking).
