Artificial Intelligence (AI) has rapidly become a part of our everyday lives. Tools such as ChatGPT can provide instant answers, explain concepts, and generate examples at our fingertips. This tool has changed the way we faced obstacles in our lives, especially how students use them for support through learning.
However, education is more than just simply obtaining information. Learning requires critical thinking, problem-solving, reflection and the ability to apply knowledge independently. While AI can be a powerful educational tool when used appropriately, concerns have emerged that excessive reliance on AI may reduce opportunities for students to develop these essential skills themselves.
As AI becomes increasingly integrated into classrooms and study environments, an important question emerges: are students using AI to support learning, or are they becoming dependent on it?
Our report today explores these aspects, where we use student survey data as well as global adoption statistics to investigate:
The analysis uses public AI adoption statistics together with responses from 557 university students regarding their experiences with AI-assisted learning.
We adopted this graph from a study by the World Bank Group, where they investigated global trends and shifts of AI in 2025. The growth of AI adoption between 2024 and 2025 was remarkable. Across all countries shown, the proportion of internet users engaging with ChatGPT increased substantially within a single year with almost a 15% minimum increase per population! This demonstrates how quickly generative AI has become part of everyday life.
To put that into perspective, Poland has a population of 37.5 million. That is an extra 5.625 million people using AI as a part of their everyday routine in just 1 year! Importantly, these figures come after AI had already become widely known, suggesting that adoption continues to accelerate rather than slow down. As AI systems become more capable, accessible and integrated into digital platforms, usage is likely to continue increasing. Students are among the groups most exposed to these technologies, making education one of the sectors most likely to experience long-term impacts from AI adoption.
Here, we can investigate how Students most commonly reported using AI for obtaining information and explaining concepts. These students were from the Vodafone Foundation, where 7 European Countries were analysed for their student’s perspectives on AI. More than half of surveyed students indicated that AI is used as an information source, while conceptual explanations were the second most common use.
However, this finding raises an important educational question. While access to information can support learning, learning itself requires students to process, evaluate and apply that information independently. Simply receiving answers does not necessarily develop understanding, memory or problem-solving ability.
The popularity of information-focused uses suggests that many students may view AI primarily as a tool for obtaining answers rather than as a partner in the learning process. This creates a challenge for educators, as students can often produce correct responses without necessarily developing the underlying skills required to solve similar problems independently.
Our next graph comes from Daffodil International University, where higher education students were surveyed regarding their use of generative AI and its cognitive, emotional and learning-related effects.
At first glance, the results appear overwhelmingly positive. Saving time received the highest average agreement score, suggesting that students view AI as an extremely effective productivity tool. Learning efficiency, reduced cognitive burden and collaborative learning also received strong ratings.
However, these results raise an important question: why do students value these benefits so highly?
Many of the highest-rated benefits are linked to reducing the amount of effort required from the learner. Reduced cognitive burden, increased efficiency and faster access to information all make academic tasks easier to complete. While this can be beneficial when used appropriately, it may also reduce opportunities for students to struggle with problems, think critically and develop solutions independently.
This does not mean AI is harmful. In fact, the results clearly show that students find AI useful and valuable. However, the popularity of benefits related to convenience and reduced effort may help explain why students become increasingly reliant on these tools as part of their learning process.
The consistently positive ratings across multiple categories help explain why AI adoption has become widespread among students and why many continue to incorporate it into their daily academic activities.
To better understand whether reliance on AI influences student perceptions, we created two summary measures.
Also from Daffodil University, we created the Benefit Score that combines several positive outcomes reported by students, including time savings, increased confidence, improved learning efficiency, learning satisfaction, collaborative learning experiences, positive emotions and reduced cognitive burden.
The Dependency Score measures the extent to which students rely on AI when making decisions, solving problems and working under pressure.
The relationship shown in the graph reveals a clear trend: students who report greater dependency on AI also tend to report substantially greater benefits from using it.
This finding is not surprising. Students who use AI frequently are more likely to experience the convenience, speed and support that these tools provide. However, the relationship also suggests something more important. The benefits students receive may reinforce continued use, creating a cycle in which greater reliance leads to greater perceived value, which in turn encourages further reliance.
Basically, the students who depend most heavily on AI are often the same students who believe they benefit the most from it.
Lastly, we stuck to the same dataset from Daffodil University to extract our dependency variable. While the previous graph demonstrated that students who depend more heavily on AI report greater benefits, this final graph explores whether increased dependency is also associated with learning-related concerns.
The Average Reduced Problem-Solving Score represents how strongly students agree with statements suggesting that reliance on AI may reduce opportunities to develop independent problem-solving ability. Higher scores indicate stronger concern that AI usage may be replacing rather than supporting parts of the thinking process.
The graph shows that students in higher dependency groups generally report higher reduced problem-solving scores. In other words, students who rely more heavily on AI are also more likely to believe that their ability to solve problems independently may be affected.
Interestingly, the color of each bubble represents the average Benefit Score. The groups reporting the strongest problem-solving concerns are often the same groups reporting the highest benefits from AI.
This creates a potential trade-off. Students appear to gain significant value from AI through improved efficiency, convenience and support. At the same time, they understand that excessive reliance may reduce opportunities to improve on critical thinking and independent problem-solving skills.
The results do not suggest that AI is damaging education. Rather, they highlight an important challenge for educators and students alike: ensuring that AI remains a learning tool that supports thinking rather than a substitute for it.
Artificial Intelligence has rapidly transitioned from a niche technology into a mainstream part of everyday life. As global adoption continues to accelerate, students are increasingly integrating AI tools into their learning processes, particularly for obtaining information, understanding concepts and completing academic tasks more efficiently.
The findings presented throughout this report show that students perceive significant benefits from AI. Saving time, improving learning efficiency and reducing cognitive burden were consistently reported as valuable outcomes. Furthermore, students who rely more heavily on AI tend to report even greater benefits, suggesting that these tools provide meaningful support in modern educational environments.
However, the results also reveal a more complex picture. The same students who report the greatest benefits are often the students most concerned about dependency on AI. While AI can provide answers instantly, education is not simply the transfer of information. Learning requires students to think critically, struggle with difficult problems, make mistakes and develop solutions independently. These are the experiences that build long-term understanding and intellectual growth.
The findings do not suggest that AI is ruining education. Rather, they suggest that the future of education depends on how AI is implemented and used. When used as a learning partner, AI has the potential to enhance understanding, improve accessibility and support student success. However, when used as a replacement for independent thinking, it risks reducing opportunities for students to develop the very skills that education is designed to develop.
The goal should not be to replace learning with AI, but to ensure that students are educated with the methods of implementing AI the right way, where it strengthens learning while preserving the critical thinking and problem-solving that remain the foundation of education.
Liu, Y., & Wang, H. (2026). Who on Earth is using generative AI? World Development, 199, 107260.
OpenAI. (2026). ChatGPT [Large language model]. https://chatgpt.com
Rahman, M. M., Munshi, T., & Kafi, M. A. A. (2026). Student awareness, cognitive strain, and emotional responses to AI assistance dataset (Version 2) [Data set]. Mendeley Data. https://doi.org/10.17632/9v9rtzch8g.2
Vodafone Foundation. (2025). AI in European schools: A European report comparing seven countries. Vodafone Foundation. https://skillsuploadjr.eu/docs/contents/AI_in_European_schools.pdf
Generative AI (ChatGPT) was used as a support tool during this project for brainstorming ideas, refining concepts, debugging code, and exploring potential approaches for developing the data story. The tool was used to assist learning rather than to generate final answers or complete the project independently.
All final analyses, visualisations, code, interpretations, conclusions, and written content submitted in this report are my own work. The overall story structure, design decisions, data selection, and final presentation were developed independently by myself.
All datasets used in this project were obtained from publicly available and credible sources, and all sources have been appropriately acknowledged and referenced