How to Use AI to Enhance Your Thesis (But Why You Should Think Twice)
(Last updated: 25 February 2025)
Since 2006, Oxbridge Essays has been the UK’s leading paid essay-writing and dissertation service
We have helped 10,000s of undergraduate, Masters and PhD students to maximise their grades in essays, dissertations, model-exam answers, applications and other materials.
If you would like a free chat about your project with one of our UK staff, then please just reach out on one of the methods below.
Producing a thesis is a lot of work. You need to develop an original topic, conduct some research, comprehend your results, and then – the most difficult task – write up the thesis and make sure it is of excellent quality. Given the complexity of this process, it is no wonder that students get stressed when their thesis is due for submission and turn to AI tools for assistance. However, while AI tools are increasingly being promoted as aids in thesis writing, they are no substitute for genuine academic expertise and should not be used to write your thesis. While AI is often promoted as a useful aid in thesis writing, it can only help at a superficial level. AI should never replace genuine academic expertise. Here’s how AI might offer some support in thesis writing, and why you should think twice before letting it take over your thesis process.
AI Can Find Sources, But Can’t Evaluate Them Critically
All dissertations and theses require a comprehensive literature review that will introduce the topic and evaluate the literature in the area. However, getting hold of the relevant literature is not the easiest task. You may need to spend days browsing academic databases to locate books and articles on the topic. AI tools such as ScholarAI can generate lists of articles, but can they critically assess the relevance, credibility, and impact of those sources? The answer is no. AI-generated bibliographies often include outdated, irrelevant, or even non-existent sources. Remember that these tools rely on algorithms, not academic judgement, so they shouldn't be employed in your thesis writing process.
AI Summarises, But It Cannot Engage in Critical Thinking
Once you have collected your literature, the next step is to comprehend, synthesise, and critique it. Some AI tools claim to summarise journal articles, but summarisation alone does not equate to critical engagement. A PhD-level literature review requires nuanced comparisons, identification of trends, and the ability to challenge existing viewpoints—areas where AI falls drastically short.
Some ChatGPT plugins, such as AskYourPDF, allow you to upload a PDF and summarise its content. You could even ask it questions to enhance your understanding of the source at hand. However, AI tools often oversimplify complex arguments, missing the subtleties that distinguish a superficial review from a scholarly one. They cannot trace how ideas have evolved over time, nor can they identify contradictions or emerging debates within the field. This kind of intellectual engagement requires subject expertise and critical insight, qualities that only experienced academics can provide.
AI Suggests Methodologies, But Lacks Research Expertise
Regardless of your area of study, your thesis will require describing a methodological approach you adopted in your research. Students often find methodologies confusing because they encompass various strategies and techniques that must be must be tailored to a specific research question.
While ChatGPT has an extensive understanding of different methodologies it lacks the ability to tailor them to the specifics of your study. It cannot justify methodological choices based on existing research, nor can it anticipate the real-world challenges of data collection and analysis. Remember that AI might recommend a generic approach, but it cannot advise on the practical implications of your chosen method, nor can it troubleshoot issues that arise during your research.
AI Processes Data, But Cannot Interpret Results Meaningfully
Within your thesis, you will report quantitative or qualitative results–depending on your research's investigation. Your results will need to be understood in the context of your thesis and such understanding will require an interpretation based on the reviewed literature and adopted methodology. ChatGPT may assist in data organisation, but it cannot contextualise results within the broader field of study or draw meaningful connections between findings and existing literature.
AI Generates Ideas, But Cannot Provide Original Insights
Your last step in the thesis-writing process will be to develop a discussion section. You will need to discuss aspects such as the strengths and limitations of your research and implications for the literature and real life. This is another section where ChatGPT can offer some insights, but they will never be original. AI-generated discussions often rely on generic, surface-level observations. They cannot engage with emerging debates, challenge established assumptions, or highlight how your findings contribute to the field.
Takeaway Message
While AI may appear helpful in thesis writing, it lacks academic judgement, subject-specific expertise, and the ability to engage in critical analysis. Remember that universities (and your field of research) require original, well-argued, and thoroughly researched work—not AI-generated content.