East China Normal University (ECNU) has received more than 700 submissions from around the world over the past three months, after calling for papers that have AI as the first author, an experimental move to challenge conventional definitions of authorship.

As AI use increasingly becomes the norm and is deeply embedded in every stage—from conception to expression and knowledge production—preserving academic authenticity and human ingenuity has become a heated topic of discussion.

Tang Zheng (Rednet.cn): No matter how much AI aids expression, genuine innovation in terms of evidence and interpretation still stems from the researcher's deep command of the field and keen insight. What matters is whether the study provides irreplaceable academic value. The focus should shift from the unanswerable question of whether the writing was entirely human-made to the more verifiable question of whether the research makes an authentic contribution.

To strengthen the verifiability of results, researchers should be required to make their data, materials and analytical processes as transparent and reproducible as possible, and to present arguments in a clear, contestable manner that can be open to critique and refutation. 

Researchers must be able to clearly articulate their understanding of the problem, a logical argument and core judgments—demonstrating in defenses, reviews and other evaluative settings that they are not merely the submitter of a text or paper, but the true "understander" and "bearer of responsibility" for the work. Otherwise, researchers may use AI to produce arguments they do not genuinely comprehend, and thus fabricate scholarly competence they never actually possess. 

ECNU's experiment tells us that rather than getting bogged down in debating whether AI can be called an "author," it is far more important to build a system that ensures what ultimately earns recognition is the researcher's genuine intellectual rigor, meticulous effort and non-transferable responsibility—regardless of the tools employed.

Bao Wanping (China Science Daily): If AI can be recognized as the first author, the implication will be all-dimensional and disruptive. It will shake the ethical foundation of the modern academic system—the sanctity of authorship. For centuries, a system of responsibility, honor and integrity has taken shape surrounding authorship, with authors responsible not only for the originality of the content, but for guaranteeing the reliability of the data and conclusions. However, if AI is put in the position of first author, troubles will arise: How can a non-human entity take on relevant responsibility when it is neither able to understand the content it outputs nor undertake legal or moral accountability for the content?

To some extent, the attempt to designate AI as the first author reveals a lag in conventional academic standards in the face of generative AI. Currently, from postgraduates to veteran researchers, AI is widely used in arranging files, data analysis and drafting documents. To ignore this fact is equal to burying one's head in the sand. ECNU's attempt is precisely a radical, straightforward confrontation with reality—one that compels the academic community to face this phenomenon head-on and to regulate it.

By its very nature, AI is a tool reliant on algorithms and data, lacking autonomous consciousness or the capacity for moral judgment. Consequently, it cannot be held responsible for the truthfulness, accuracy or ethical compliance of the content it generates, nor can it shoulder the consequences arising from academic flaws or erroneous conclusions. Therefore, scholars who collaborate with AI in their research must bear ultimate academic and ethical responsibility for the papers produced. In this collaboration, AI functions as a tool—something akin to a super "scribe" or "calculator." The core value of the human researcher, on the other hand, lies in profound insight, rigorous reasoning, adherence to academic integrity and the holistic control over the direction and quality of the research. BR

Copyedited by G.P. Wilson 

Comments to panxiaoqiao@cicgamericas.com