An AI summary can tell you what a paper says (beware of common errors like hallucinations and pulling in information from other sources without disclosure). It does not tell you whether the argument is robust. Before you cite a paper through a summary, run these five checks. Each maps to one element of the CERIC Method.
Claim: is the summary’s claim the paper’s claim?
Find the sentence in the abstract or conclusion where the authors state their main finding. Compare it with the summary. Summaries often strengthen hedged claims: “is associated with” becomes “causes”.
Evidence: can you point to it?
Locate the figure or table that carries the main claim. If you cannot find it, you are not ready to cite the paper.
Reasoning: does the evidence support the claim?
Ask what else could explain the result. Check whether the authors addressed it.
Implications: what do the authors say follows?
Separate what the study shows from what the authors hope it means. Summaries tend to blend the two.
Context: why was the study done?
Find the gap or rationale the authors give in the introduction. Summaries usually drop it, and it tells you what question the paper was built to answer and how it fits the work that came before.
Five checks take about fifteen minutes with practice. Try them with the one-page cheat sheet, or learn the method step by step in the free introductory course. The book, How to Critically Read the Scientific Research Literature, covers each element in depth.