Dear Colleagues,
We are living in interesting times.
In addition to the concern of data integrity issues that we are all facing, there is a new concern about using AI tools to review papers. Some science journals have developed policies against the use of AI tools to review a paper submitted to a journal. Springer publishers have just developed one.
Knowing one size does not fit all, I have asked our EICs of all 17 journals to work with their editorial board members to develop a policy that is more suitable for their journals. With the help of the Managing Editor of each journal, such a policy will be posted on the journal website. Just keep an eye out for this policy. If you have suggestions and comments, please contact the corresponding EIC.
For example, Alice Smith, EIC of INFORMS Journal of Computing has worked with her board to develop the following policy, and the Managing Editor Kelly has posted this policy on IJOC websites: author guidelines / announcement / quick link. Now other journal EICs are working with their boards to customize their policies for their journals.
Ultimately, I need your help. Once you agree to review a manuscript, please review it yourself without uploading the paper and ask the AI tool to generate a report for legitimate reasons as stated in the following policy. As INFORMS members, we should show our commitment and our respect to our colleagues for their research work.
Thank you. With your help, INFORMS can continue to thrive.
Chris Tang
INFORMS VP of Publications
INFORMS Journal on Computing
Reviewing and Generative Artificial Intelligence Policy
August 2023
Reviewing a technical manuscript implies responsibilities that can only be attributed to humans. The critical thinking and assessment required for peer-review are outside the scope of generative AI and AI-assisted technologies, and there is a risk that the technology will generate incorrect, incomplete, or biased outcomes. These considerations, together with the principle that submitted manuscripts are to be treated as confidential documents, results in the IJOC Generative AI policies for reviewers and editors:
Generative AI should not be used to assist in the review, evaluation, or decision-making process of a manuscript. Some specific examples are below but these are not comprehensive:
• Reviewers or editors should not upload the manuscript or any part of it into a Generative AI tool, as there is no guarantee of where materials are being sent, saved, or viewed, or how they may be used in the future to train the model, and this may violate author confidentiality.
• This confidentiality requirement extends to the peer review report and any other communication about the manuscript, such as the notification or decision letters, as they may also contain confidential information about the manuscript and/or the authors. For this reason, they should not be uploaded into a Generative AI tool, even if it is just for the purpose of improving language and readability.
If an editor or reviewer suspects Generative AI has been used in the paper under review which has resulted in improper aspects such as bogus references, nonsensible reasoning, or misinformation that person should report the matter to the editor in charge, normally the area editor.
------------------------------
Christopher Tang
Distinguished Professor
ucla
Los Angeles CA
------------------------------