Large language models (LLMs) are reshaping academic research, but their use requires careful oversight, writes Ali Shiri (University of Alberta). Shiri highlights how LLM tools like ChatGPT and Gemini can support nearly every stage of the research process and foster interdisciplinary collaboration, but raises concerns around bias, privacy, and citation integrity. He stresses the “urgent need to develop AI literacy training tailored for academic researchers” so they can distinguish between “AI research assistants” and emerging “deep research” agents that conduct multi-step analyses across disciplines. Shiri concludes by pointing readers toward the guidelines being published by government and postsecondary sources, as these serve as valuable steps toward the responsible, ethical integration of LLMs in the research process.
