SCIENCE · VERIFIED DEVELOPMENT
AI Chatbots Widely Used, but Nearly Half of Their News Answers Are Wrong, Study Finds
WHY IT MATTERS
Because students increasingly use AI for research, unchecked inaccuracies can compromise academic work and spread misinformation, making critical evaluation skills essential.
What happened
AI chatbots are now a daily companion for almost a quarter of U. S. adults, yet a recent study shows that 48 % of the news‑related answers they provide contain errors.
Librarians warn that students who rely on these tools for academic research must treat AI output skeptically. They recommend first selecting a tool that matches the task—ChatGPT for overviews, specialized services like Primo Research Assistant or Scopus AI for peer‑reviewed literature. Next, ask follow‑up questions and request source citations.
Then verify each claim against the original material, checking author expertise and context. Finally, remain alert to hidden biases and missing perspectives, and always cite the AI when quoting it. These steps help safeguard research integrity in an era of rapid AI adoption.
PRIMARY SOURCES
AI can help with academic research but doesn’t always get it right – here are some tips to check the information you receive
The Conversation US · Allison Faix, Librarian and Instruction Coordinator, Coastal Carolina University · CC BY-ND; link/attribution intake only—no edited republication
CORRECTIONS & UPDATES
- Revision 1 · Initial ingestion · Oct 7, 2026, 1:30 PM
- Revision 2 · Source update detected · Oct 7, 2026, 1:30 PM