Scientists are failing to disclose their use of AI despite journal mandates, finds study

Scientists are failing to disclose their use of AI despite journal mandates, finds study

A recent study has revealed a significant gap between the increasing use of artificial intelligence (AI) tools in academic writing and the number of researchers who are disclosing their reliance on such technologies, even when journals have policies requiring such disclosures. The analysis, conducted by data scientist Yi Bu from Peking University and colleagues, examined over 5.2 million papers published in 5,000 different journals between 2021 and 2025, with a focus on the use of AI in scientific writing across all disciplines, particularly in physics.

The study found that the use of AI writing tools has surged dramatically since 2023, with a notable increase observed across all scientific fields. This rise in AI usage has prompted many journals to introduce policies requiring authors to disclose the use of AI-writing tools. As of now, 70% of journals have adopted such policies, with IOP Publishing, which publishes Physics World, being one of them. IOP Publishing's policy encourages authors to be transparent about their use of generative AI tools in both the research and the drafting of the manuscript.

Despite these efforts, the study uncovered a concerning "transparency gap." A full-text analysis of 75,000 papers published since 2023 revealed that only 76 articles (approximately 0.1% of the total) explicitly disclosed the use of AI writing tools. This indicates that researchers are not adhering to the disclosure requirements set by their journals. Furthermore, the study found no significant difference in the use of AI between journals that have disclosure policies and those that do not, suggesting that these requirements are being largely ignored.

The researchers' findings highlight the need for stricter enforcement of AI disclosure policies and the development of more effective frameworks to ensure transparency in academic research. As AI continues to play an increasingly significant role in scientific writing, it is crucial that the scientific community remains vigilant about the potential implications of undisclosed AI usage. This includes the risk of biased or inaccurate research findings, as well as concerns about authorship and the integrity of the scientific process.

In light of these findings, the study calls for a reevaluation of current AI disclosure policies and the implementation of more robust mechanisms to ensure compliance. This may involve stricter penalties for non-disclosure, as well as the development of standardized protocols for AI usage in research. Additionally, the study emphasizes the importance of fostering a culture of transparency within the scientific community, where researchers are encouraged to openly discuss and acknowledge the role of AI in their work.

The growing reliance on AI in academic writing raises important questions about the future of scientific research and the role of technology in shaping the field. As AI continues to evolve and become more sophisticated, it is essential that the scientific community remains proactive in addressing the challenges posed by its integration into the research process. By doing so, the community can ensure that the benefits of AI are harnessed while minimizing the risks associated with its misuse or undisclosed application.

In conclusion, the study's findings underscore the urgent need for greater transparency in the use of AI in scientific writing. While many journals have implemented policies requiring AI disclosure, the reality on the ground shows that these requirements are not being met. To address this issue, the scientific community must work together to enforce existing policies more effectively and develop new frameworks that promote transparency and accountability in the use of AI in research. Only through such efforts can the potential pitfalls of AI in academic writing be mitigated, and the integrity of the scientific process be preserved.