AI, Plagiarism, and Real Benefits: Insights from the 19th Math Modeling Conference

Focus on AI Usage

AI is everywhere, whether or not it appears in the competition theme, and its impact on modeling contests, research, and industry is widely discussed.

When asked how the committee views AI, Secretary‑General Prof. Xie Jin‑xing stated that the competition does not oppose AI use, emphasizing that AI is an advanced productive force and should not be rejected.

However, contest designers will consider how to set problems that highlight students' innovation and creativity.

Deputy Secretary‑General Zhang Wen‑bo added that AI can be used but must adhere to academic norms, and any AI‑generated content should be clearly acknowledged, e.g., "honor to AI," rather than presented as the student's own work.

Plagiarism Rate Concerns

Last year many teams missed awards due to high similarity scores. Prof. Xie explained that similarity rates are objective data, but their interpretation should be tailored to each region; a high rate does not automatically indicate academic misconduct.

Nevertheless, a similarity above 25% is generally frowned upon, and papers exceeding this threshold are often excluded from national evaluation.

Similarity data come from two sources: public checks on CNKI and the competition’s own database of submitted papers. Overlap can also arise when teams copy problem statements during the “problem restatement” section.

AI‑generated content can increase similarity rates, especially when many participants use identical prompts, leading to similar outputs.

Importance of Accurate Results

National competition judges, led by Prof. Cai Zhi‑jie, stressed that result accuracy is always important, and teams should include members with strong programming skills. While accuracy matters, model innovation is equally crucial for comprehensive evaluation.

There is no fixed scoring rubric; judges assess modeling process, result accuracy, innovation, and writing standards.

Beyond Awards: The True Value of Participation

Although many aim for awards, only about 1,500 out of 60,000 teams receive national recognition, making award expectations unrealistic for most.

Prof. Xie highlighted that the competition’s motto is "One participation, lifelong benefit," not "One award, lifelong benefit." Participants gain integrated training that combines mathematics, computing, domain knowledge, and real‑world problem solving.

The contest also hones self‑management, autonomous learning, teamwork, communication, pressure handling, and conflict resolution—skills often beyond classroom teaching.

Many teachers reported that participation led to undergraduate publications, stronger graduate school interviews, and deeper subject understanding.

If you have further questions, feel free to leave a comment in the 留言区 for answers from the community.