Chemists and editorial boards of scientific journals have proposed rules for the responsible use of AI in preparing articles. This was reported to TASS by the Ministry of Education and Science. The use of AI should not be a reason for rejecting a manuscript.
The VALENCE rules include seven principles. The author is obliged to verify facts and references, openly indicate significant application of the model, limit its role to auxiliary tasks, carefully edit the result, not list AI as a co-author, correctly cite sources, and protect confidentiality. AI is permissible for translation, editing, and text structuring, but scientific conclusions and responsibility remain with the human.
For authors, a declaration model was proposed: to state the fact and purpose of AI use, confirm human verification and approval of the text. AI cannot be an author — it does not bear scientific or legal responsibility. Among the co-authors of the rules are 20 academicians of the Russian Academy of Sciences, two vice-presidents of the Russian Academy of Sciences, and chief editors of chemical journals.
Published in the journal "Uspekhi Khimii" (Russian Chemical Reviews), including in English.
Read more on the topic:
- AI Law Comes into Force: Sovereign and National Models Defined
- Bringing AI out of the "gray area": Russian universities and research institutes are developing regulations for the use of neural networks
- Neural networks given age limit: Ministry of Education defines rules for schoolchildren