Stormy times

Abstract

Much more rapidly than anticipated, we have observed the submission of manuscripts to the Journal of the Foot and Ankle that have been predominantly produced with the assistance of Artificial Intelligence (AI).

This observation stems primarily from the diligent work of our editorial board, which is committed to providing constructive and collaborative reviews rather than merely selective evaluations. This editorial philosophy obliges each reviewer to engage deeply with the manuscripts, working alongside the authors to improve not only the structural formalities of a scientific paper but also the clarity and precision of the language. In doing so, reviewers decisively influence the quality of the final product, thereby enhancing the dissemination of ideas and results, and facilitating the incorporation of new knowledge into daily practice.

However, this task has increasingly become exhausting — and at times nearly unmanageable — when confronted with lengthy, complex texts that are clearly inflated by the use of AI. Such manuscripts are often riddled with thematic confusion, inconsistent insertions attempting to establish non-existent correlations, and a troubling duplication — or even multiplication — of papers on similar topics that overlap in an abnormal and unacceptable manner.

The same phenomenon has been identified in high-volume preprint platforms, where the participation of so-called “paper mills” and AI-based tools has significantly increased the proportion of LLM-generated manuscripts, reaching figures as high as 10% in the biological sciences and 22% in computer science.

The “symptoms” of AI-driven manuscripts may be summarized as follows:

1. Overreliance on technology – texts produced without the rigor of conventional scientific methodology;
2. Stereotyped writing – instead of advancing knowledge, such texts contribute to increased “noise” within scientific fields;
3. Authorship opacity – lack of transparency regarding the role of AI in manuscript production;
4. Fabricated claims – due to AI “hallucinations” that introduce false data and references;
5. Rapid dissemination of misinformation – amplified by modern preprint publication systems;
6. An “arms race” between AI authors and AI detectors – which may generate false positives and unfairly penalize legitimate authors.

The central concern, therefore, lies in the undeniable risk of disseminating low-quality articles that circulate misleading or false information, thereby severely undermining the trustworthiness of scientific output.

In our ongoing efforts toward the indexation of the Journal of the Foot and Ankle, we must now contend with this additional challenge. Stricter differentiation between authentic manuscripts and those generated by AI will be indispensable, as it represents the only viable means of safeguarding the integrity — and the very survival — of this journal.