Штучний інтелект і академічна доброчесність: запитання без відповідей

This study examines how leading international academic publishers have developed and evolved their policies regarding AI use in scholarly publishing. Researchers analyzed websites of five major academic publishers (identified through AI consultation) to investigate their approaches to AI regulation. The study conducted two analyses separated by two years to track the evolution of policy on AI use. The results showed that all publishers agreed that AI should never generate text published under human authorship, essentially contradicting generative AI's primary designed purpose. A comparative analysis revealed fundamental shifts in publisher approaches. Publishers developed detailed guidelines for authors, reviewers, and editors on how to use and check AI in research. The research demonstrates that publisher policies cannot be formulated due to AI's rapid evolution. Constant policy review and adaptation are necessary as technology capabilities continue expanding. Despite policy evolution, AI tools cannot serve as article authors or receive authorship credit. It maintains human accountability in scholarly attribution and publishing standards. The study emphasizes that technological detection methods alone cannot address fundamental integrity issues. No software can replace the moral responsibility and ethical awareness that researchers must personally follow. The research concludes that AI tools like ChatGPT should function as research aids rather than manuscript substitutes. The future of academic integrity depends on balancing technological assistance with human responsibility, ethical awareness, and continuous policy adaptation in scholarly work. Key words: artificial intelligence, academic integrity, confidentiality, publishing policy, ethical norms.