Original title: Testing Our Foundations: Citation Trends, Errors, and Emerging Hallucinations in the Computing Education Literature
Paul Denny, Gweneth Barbre, Musa Blake, Yan Cathy Hua +4 · September 2026
Researchers need real citations to verify facts and give credit. Now, people use large language models, or AI tools that write text, which can fabricate realistic fake references. This paper checked how common these fake citations are in computing education research. The team inspected papers from the ACM Digital LibraryA large online archive of computer science papers and conference proceedings. and looked closely at bibliographies to spot errors and entirely invented citations. These fake references showed fake titles, impossible page counts, and wrong author names. The researchers found fake citations across five computing education conference venues in 2025. At the Technical Symposium, verified fake references climbed from 3 in 2025 to 17 in 2026. While currently rare, this rising problem poses an integrity risk that researchers should address.
You are writing an essay, and you borrow a quote from a friend. That friend says it came from a book they read. You trust them, so you write the page number down. But the book doesn't exist, and the page number is impossible. Someone later tries to find that quote and hits a dead end. In this paper, the researchers are checking thousands of essays to find all the quotes that came from untrustworthy friends. They found that the number of these fake references is growing, especially in recent conference papers.
Every researcher builds on past work, and fake citations poison that foundation. If you can't trust a reference list, you can't verify claims or reproduce studies. This paper shows the problem is already inside a major education conference, and it's growing every year.
The abstract focuses on computing education venues and uses only ACM Digital Library data. It does not assess the prevalence of hallucinated references in other fields or databases, nor does it measure the impact of these errors on downstream research. The manual classification may miss hallucinated references that are too subtle to detect by eye.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Testing Our Foundations: Citation Trends, Errors, and Emerging Hallucinations in the Computing Education Literature — Paul Denny, Gweneth Barbre, Musa Blake et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.