Artificial Intelligence (AI)

Generative Artificial Intelligence (GenAI) has had a significant impact in academic publishing. As we all try to come to grips with a phenomenon that has taken the world by storm, some researchers have even tried to list ChatGPT as their co-author.

AI should be used to enhance, not compromise, the integrity and reproducibility of research findings, with a strong emphasis on transparency and human accountability...

• Transparency
  • maintain openness about the methods and processes used
• Human Accountability
  • ensure that humans are ultimately responsible for the research outcomes.

Numerous policies and guidelines have been put into place. If you are thinking of publishing, please read the author guidelines for each publisher very carefully.

Policies and Guidelines set by Publishers and Professional Organisations

The Committee on Publication Ethics (COPE) issued a statement on authorship and AI, affirming their belief that AI cannot be considered an author of a publication...

• Authorship
  • AI tools cannot be listed as authors because they cannot take responsibility for the work or manage conflicts of interest and copyright agreements
• Disclosure
  • authors must disclose how AI tools were used in their research and are fully responsible for the content, including parts produced by AI tools
• Publication Ethics
  • authors are liable for any breach of publication ethics, even for content generated by AI tools.

Whether publishing, reviewing or editing, it is essential to check the publisher’s policies on AI use. For example, Elsevier Publishing Ethics (policies on AI use can be found under Duties of Editors, Duties of Reviewers, and Duties of Authors).

The use of generative AI and AI-assisted technologies in scientific writing...

• Human Oversight
  • AI should only be used to improve readability and language, with human oversight and control. Authors must review and edit AI-generated content
• Disclosure
  • authors must disclose the use of AI in their manuscripts, and this will be stated in the published work
• Authorship
  • AI cannot be listed as an author or co-author. Authorship responsibilities can only be attributed to humans.

The use of generative AI and AI-assisted tools in figures, images and artwork...

• Image Adjustments
  • Generative AI or AI-assisted tools are not allowed to create or alter images in submitted manuscripts, except for brightness, contrast, or colour adjustments
• Research Design
  • AI tools can be used if they are part of the research design or methods, with detailed descriptions and proper attribution
• Graphical Abstracts
  • Generative AI is not permitted for graphical abstracts
• Cover Art
  • Generative AI may be used for cover art with prior permission (from the journal editor and publisher) and proper rights clearance.

Rigorous Peer Review Processes

• Enhanced Review Processes
  • journal publishers are incorporating AI tools to assist in detecting plagiarism, data fabrication, and image manipulation. However, human oversight remains crucial to verify the results
• Plagiarism Detection Software
  • journal publishers are employing advanced software tools to detect plagiarism and ensure that the data and findings presented are original and not misrepresented
• Ethical Screening
  • submissions are subjected to ethical screening processes to ensure compliance with publication standards and the integrity of the research, including proper data collection, analysis, and reporting practices.

More information can be found on this page from the Elsevier website at https://www.elsevier.com/about/policies-and-standards/publishing-ethics#3-duties-of-reviewers

Data Analytics

AI can assist in processing large datasets, identifying patterns, and generating visualisations that help researchers interpret their data more effectively. To ensure data integrity and transparency, it is crucial to...

• Data Quality
  • use high-quality, representative datasets to train AI models and be transparent about the sources and limitations of the data used.
• Reproducibility
  • provide sufficient detail about the AI methods and data so that other researchers can reproduce the results. Clearly explain the AI methods used, including algorithms, training processes, and validation techniques.

When you are making plans to implement AI in your research, you should ensure that it is included in your data management plan (DMP) - see the SIT LibGuide Research Data Management. This helps in effectively handling your data and maintaining transparency in how the AI is used and its impact on the research process.