With powerful AI tools now capable of generating literature reviews from a single prompt, it is tempting to answer this question with a quick yes or no. But before we rush to a binary conclusion, it is worth pausing to ask a more fundamental question: what is a literature review, and why do we do it in the first place?
When you begin a Ph.D. program (whether in India or abroad) the journey usually starts with coursework. These courses are not mere formalities; they provide the intellectual foundation on which your research career is built. Think of this process as building a house. Where would you start? If the first thing you consider is the room layout or the position of windows, you are clearly missing something essential. Any sensible construction begins with the foundation.
A strong foundation determines how much load a building can carry, how many floors it can support, and whether it can withstand natural stresses such as floods or earthquakes. In the same way, rigorous coursework equips a researcher with the conceptual strength needed to handle complex scientific problems.
But a foundation alone does not make a home livable. Once the foundation is laid, we move on to design and construction. Similarly, once coursework is completed, a researcher begins to shape a research plan. How is this done? By looking at existing structures: houses we have lived in, designs we admire, and solutions that have worked elsewhere, and by consulting experts who understand the constraints and possibilities.
This is where the literature review enters the picture. Surveying existing work helps us understand what has already been built, what design choices were made, which approaches succeeded or failed, and why. By carefully examining prior studies, their methods, assumptions, and outcomes, we learn what is feasible, what is promising, and where genuine gaps exist. This process does more than summarize knowledge, it gradually transforms a student into an expert capable of asking new and meaningful questions.
Now, can AI replace this process?
One of my research scholars recently shared an experience of using an AI tool to summarize research on a specific topic. The output was impressive—clear, structured, and even presented in neat tables. As a tool for organizing existing information, AI performs remarkably well.
However, what the AI could not do was arguably the most important part: it could not chart a future course of action. It could not weigh subtle trade-offs, question underlying assumptions, or anticipate how a field might evolve under new constraints or discoveries. Research is not merely about compiling what is known; it is about developing judgment—something that emerges from deep engagement, not automated synthesis.
AI can assist us in analyzing the past and the present. But extrapolating this knowledge into the future, identifying what should be done next, requires human intuition, creativity, and critical thinking. Whether AI will ever fully acquire this capability remains an open question. For now, review articles remain not just relevant, but essential—as exercises in thinking, not just in summarizing.
Only time will tell how far AI can go. Until then, the literature review remains the intellectual blueprint on which meaningful research is built.