LSE - Small Logo
LSE - Small Logo

Thorsten Fröhlich

June 27th, 2025

How students should write a thesis in the AI age

1 comment | 21 shares

Estimated reading time: 5 minutes

Thorsten Fröhlich

June 27th, 2025

How students should write a thesis in the AI age

1 comment | 21 shares

Estimated reading time: 5 minutes

As AI tools reshape academic writing, students are producing technically proficient theses without truly understanding their work. Thorsten Fröhlich explores how rigid approaches to writing a thesis are stifling critical thinking, and why educators must urgently rethink how research is taught and assessed.


We are facing a quiet but profound crisis. Students are increasingly equipped with powerful AI tools, like ChatGPT, yet many are losing the crucial ability to think critically and creatively. If you’re involved in supervising student research or guiding academic writing, you’ve likely observed this shift. Thesis-writing, despite technological advancements, often remains rigid, formulaic and disconnected from meaningful thought.

Let’s pause and ask ourselves: have we fundamentally misunderstood what thesis-writing should achieve? What if the core goal isn’t ticking boxes or applying techniques but nurturing deep, reflective thinking?

My recent research is designed to help educators foster meaningful intellectual growth through thesis-writing in the age of generative AI.

Is AI a helpful tool or a barrier to thinking?

AI-driven language models have dramatically simplified thesis writing, providing ideas, streamlining complex tasks and breaking language barriers. Yet, these same tools can become barriers to genuine understanding.

A recent student confidently cited AI-generated content that, although convincing on the surface, contained significant inaccuracies.

I’ve observed students submitting sections produced by AI without fully grasping their content. For instance, a recent student confidently cited AI-generated content that, although convincing on the surface, contained significant inaccuracies. This underscores the urgency of promoting critical reflection and active thinking, rather than passive reliance on AI-generated information.

Are we losing the art of asking good questions?

This challenge predates AI, but generative tools may be accelerating the problem. Recent research demonstrates a significant negative correlation between frequent AI tool usage and critical thinking abilities. When students can quickly generate plausible-sounding research frameworks, the temptation to skip the crucial step of questioning whether these approaches truly fit their specific research context becomes even stronger.

Higher confidence in AI is associated with reduced critical thinking effort.

Moreover, studies show that higher confidence in AI is associated with reduced critical thinking effort, suggesting that easy access to AI-generated solutions may be discouraging the fundamental skill of asking probing questions about methodological appropriateness. Encouraging students to challenge and rigorously question their methodological choices is essential for fostering deeper insight and innovation. Indeed, there is now a greater need to break free from methodological orthodoxy.

Some might, instead, argue that methodological consistency ensures research rigour and comparability. While consistency is valuable, strict adherence can inhibit originality and deeper understanding. I’ve seen students hesitating to experiment with new or interdisciplinary methods, fearful of criticism or rejection. Encouraging thoughtful methodological flexibility can lead to more prosperous, innovative research outcomes.

A call to action for educators

Imagine if thesis-writing were seen not merely as a graduation requirement but as a valuable exercise in deep reasoning. Rather than mechanically fulfilling academic standards, students would thoughtfully engage with assumptions, justify their choices rigorously and develop coherent arguments. For instance, students who embraced this approach in my seminars created notably more insightful theses and demonstrated more potent reasoning abilities.

During supervision meetings, begin by regularly challenging students to justify their methodological choices and assumptions explicitly. Assessment criteria should reward clarity, originality and depth of reasoning rather than conformity.

Now is the time for supervisors and educators to actively prioritise critical thinking and reasoning skills. During supervision meetings, begin by regularly challenging students to justify their methodological choices and assumptions explicitly. Assessment criteria should reward clarity, originality and depth of reasoning rather than conformity. By fostering reflective thinking over procedural compliance, we better equip students to navigate complex and uncertain future challenges.

Try this: start your next supervision meeting by asking, “Why did you choose this method over alternatives?” It’s a simple but powerful prompt that can shift the conversation from compliance to reasoning.

The AI revolution in academic writing presents both opportunity and risk. By shifting our focus from procedural compliance to deep reasoning, we can ensure that these powerful tools enhance rather than replace critical thinking. The future of thesis-writing lies not in choosing between human intellect and artificial intelligence, but in using AI to free students to engage more deeply with the fundamental questions that drive meaningful research.

***

These ideas and recommendations emerge from extensive research and practical experience detailed in The Art of Thesis Writing: From Understanding Research to Writing Success (2nd Edition), a resource designed to help educators foster meaningful intellectual growth through thesis writing.


The content generated on this blog is for information purposes only. This Article gives the views and opinions of the authors and does not reflect the views and opinions of the Impact of Social Science blog (the blog), nor of the London School of Economics and Political Science. Please review our comments policy if you have any concerns on posting a comment below.

Featured image credit: Nadia Piet & Archival Images of AI + AIxDESIGN, Better Images of AI, (CC-BY 4.0).


About the author

Thorsten Fröhlich

Thorsten Fröhlich is an IT Management and Big Data Professor at the IU International University of Applied Sciences. His expertise includes applied artificial intelligence, machine learning, big data, IT management, and digital marketing. His research focuses on the practical application of AI in education, ethics, and natural language processing (NLP). As a recognised expert, he lectures on AI and ethics in education in collaboration with UNESCO, among others. He recently published a comprehensive compendium on scientific writing with M.C. Hemmer. After studying chemistry at the University of Cologne, he completed his doctorate in physical chemistry in 1992. He has been an entrepreneur since 1987 and has founded several successful companies in software development, IT management, and digital marketing for the life science industry. As a business angel, he supports student start-ups and promotes innovative technology projects.

Posted In: AI Data and Society

1 Comments