AI Policy

EAI Artificial Intelligence Policy 

Artificial Intelligence (AI) technologies have been evolving at a rapid pace and they have become part and parcel of various stages of the research and publication processes, changing how we approach academic knowledge production. While AI technologies offer a variety of research and publication process enhancements, their presence in these processes also raises concerns about their ethical and transparent use. 

Authorship

EAI follows the COPE position on authorship and AI tools which does not recognize AI tools as authors of academic texts. Since AI tools cannot take accountability for the work submitted and ensure the integrity of their data and results, they do not meet the EAI authorship criteria. Furthermore, given their status as non-legal entities, AI tools cannot handle copyright and license agreements nor claim the existence or lack of conflict of interest. AI tools thus cannot be listed as authors, co-authors, or contributors to a manuscript, and neither can they be cited as such. The correct and transparent use of the AI by authors tools is discussed in the next section of the policy. 

Transparency in the use of AI

To ensure academic honesty and the integrity of the publishing process, EAI requires that authors disclose the use of AI technology which goes beyond basic support of human writing processes. Basic AI writing assistance includes actions such as correction of grammar or improvement of the clarity and fluency of the text, and does not need to be declared by an author, but is recommended.

If AI technology is used by authors in a more substantial manner than this, authors are required to disclose and elaborate the precise use of AI tools in their research and writing process in the methodology section (or another suitable section) of the manuscript. The author’s use of AI technologies will be evaluated by the editorial team of the journal. In manuscripts where substantial use of AI tools has not been declared, EAI reserves the right to remove these texts from the editorial process. 

Machine translations

While we recognize the challenges authors face when writing academic texts in their non-native languages, we encourage authors not to use machine translations to translate entire texts without competent human oversight. Without thorough editing by a qualified human translator, these machine translations often generate ‘tortured phrases’, which distort the original meaning, making the text incomprehensible for the reader. If AI is used for translation, it should be declared by the author in the submitted text. 

Use of generative AI figures, images, and videos

The use of AI generated images and manipulation of images using AI technology is not permitted except for the following instances when:

  • Generation of AI images is integral to the research design of the manuscript and author provides a detailed elaboration of this in the methodology section

  • AI generated images and videos are used as references in a paper focused on use of AI

When AI generated images and videos are used by authors, they must be clearly labeled as such. The editorial team will review these exceptions on a case-by-case basis.

Peer review 

To preserve confidentiality and integrity of the peer review process, reviewers cannot upload the manuscript they were assigned into an AI software. The evaluation and assessment of the manuscript should be the sole responsibility of a human reviewer and such tasks are outside of scope of AI technology, which cannot take accountability for such decisions. Furthermore, uploading the assigned manuscript into an AI software could disclose proprietary and sensitive information, thus constituting a violation of intellectual property law and breach of data privacy rights.