Skip to content
Submit incident
AII-1001-46713
Under investigationA regulator, court or police has opened a formal inquiry and reached no finding yet.

University of Hong Kong Paper Contains AI-Generated Fabricated References

A University of Hong Kong doctoral student's published paper on Hong Kong's fertility rate contained dozens of fabricated academic references reportedly created by an AI system used to organize citations. The corresponding author, an associate dean, apologized and acknowledged his responsibility for failing to verify the references, though he stated the paper's core research findings were sound and had passed peer review.

September 2025 · Bai Yiming · Reliability and hallucination
September 2025
AII-1001-467132 sources ↓
Quick facts
Category
Reliability and hallucination
Country
Hong Kong
Source
香港01
Date
September 2025
Drafted by
Raidu Incident Detection

What happened

A paper titled 'Forty years of fertility transition in Hong Kong' published in September 2025 in the journal China Population and Development Studies by authors including University of Hong Kong doctoral student Bai Yiming and Associate Dean of the Faculty of Social Sciences Paul Yip contained numerous references to non-existent academic papers.

Of 61 citations in the paper, 22 articles with Google Scholar links could not be found, and 5 articles with DOI links showed 'DOI Not Found'. One non-existent reference was attributed to Yip as author.

Yip stated that the doctoral student used AI to assist with organizing references but failed to verify the content. Yip acknowledged responsibility as corresponding author and apologized but stressed the paper had passed two peer reviews and contained no fabricated research content.

Reported impact

The incident raised concerns about the integrity of the research and the proper use of AI in academic work, with observers questioning the trustworthiness of the study.

Index analysis

The paper's core conclusions about Hong Kong's fertility rate being driven by declining marriage rates remained valid according to the journal publisher Springer Nature, which stated the errors did not affect the article's central conclusions, empirical data validity, or theoretical framework completeness.

Governance controls

The University of Hong Kong stated it has strict guidelines for the use of artificial intelligence in research and has opened a review of the relevant staff members. The university stated that if research handling errors are found, it will follow procedures to address them.

Yip stated he will personally verify every citation in the corrected version of the paper. The journal agreed to publish a corrected version within days and will issue a notice about the incident. Yip required the student to complete a course on proper AI use.

Classification

Category
Reliability and hallucination
Industry
Education and training
Country
Hong Kong
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Reputational, Public interest
Failure modes
Hallucination, Fabricated citation
Consequences
None recorded, Public apology
AI systems
AI referencing/citation tool

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

香港01
港大論文引虛構文獻 葉兆輝致歉:博士生用AI沒檢查 非誠信問題
9 November 2025
Yahoo News
港大論文爆「AI 幻覺」爭議 涉引虛假文獻 葉兆輝稱博士生未檢查 港大:展開核查|Yahoo
10 November 2025