Researching with AI
Research has changed fundamentally. Knowing how to use AI as a research tool — and when not to — is now a core information literacy skill.
AI versus search engines: different tools
A search engine retrieves documents that exist; you choose which to read and trust. An AI synthesises across many sources and gives you a response — but that response may be inaccurate, outdated, or missing important perspectives. Use both, for different purposes.
Good uses of AI in research
- Getting oriented. When starting on an unfamiliar topic, ask AI for a broad overview and the key debates. Treat this as a map, not the territory.
- Identifying terminology. Ask what terms experts use to discuss this topic — then use those terms in database and search engine searches to find primary sources.
- Summarising long sources. Paste in a paper or report and ask for a summary of its main argument and evidence.
- Checking your understanding. After reading a source, tell AI what you think it said and ask if you have it right.
Where AI fails in research
AI will sometimes cite studies that do not exist, attribute quotes to people who never said them, and present outdated information as current. Never use an AI-provided citation without verifying that the source actually exists and says what the AI claims it says.
Evaluating sources
The SIFT method (Stop, Investigate the source, Find better coverage, Trace claims to origins) is a useful framework for evaluating any information — whether it comes from AI, social media, or traditional media.
AI is a powerful research accelerator for people who already know how to evaluate sources. In the hands of someone who does not, it is an efficient way to produce confidently wrong conclusions.