Review Process

IEEE Transactions on Artificial Intelligence


Paper Handling Process

For a detailed discussion on the review process and rules for IEEE publications, please refer to the IEEE Publication Services and Products Board Operations Manual ( The current document offers a quick guide to the paper handling process for authors to consider when writing their manuscript, and for reviewers to use as a guide to provide their review.

The formal scope for TAI is

The IEEE Transactions on artificial intelligence (TAI) is a multidisciplinary journal publishing papers on theories and methodologies of Artificial Intelligence. Applications of Artificial Intelligence are also considered.


Editorial Process

From the time a paper is received to the time a final decision is made, the workflow associated with each paper is explained below including submission instructions.


Before Submission

Authors should read this document to familiarize themselves with the expectations of the journal. The author submitting a manuscript is expected to have an account on IEEE Author Portal 


Editorial Assistant Compliance and Plagiarism Checking

A submitted paper is first checked with the Editorial Assistant to ensure

  • There is no corrupted file during the loading of the paper
  • Double-anonymized
  • Appropriate Formatting
  • Compliance with page limit
  • Compliance with Ethics for papers involving human and/or animal experiments

A manuscript that fails any of the above, will be returned to authors.

If the manuscript passes the first checklist above, the Editorial Assistant will run a plagiarism report. The maximum allowable similarity index for a manuscript is 20% assuming that statements are quoted and cited. Manuscripts that exceed this threshold will receive a straight reject.


Editor in Chief

The Editor in Chief may reject a manuscript without sending it for review if:

  • The author(s) have not followed the IEEE guidelines for style.
  • The author(s) have violated IEEE Policies.
  • The manuscript is incomprehensible (in other words, so poorly written that it is unreadable).
  • The subject and contents of the article do not meet the scope of the periodical or a specific special issue, or the abstract, introduction and/or conclusion are inaccessible to the journal’s wide readership pool.
  • The manuscript raises ethical concerns for manuscripts involving human and/or animal participation.
  • The manuscript does not meet the very minumum techncal grounds for the journal.

If a manuscript passes the above, the Editor in Chief will assign an appropriate Associate Editor to handle the manuscript.


Associate Editor

Manuscripts will be assigned 3-4 reviewers. Decisions can be made on manuscripts based on two reviews only if and when the Associate Editor sees this appropriate. For example, the Associate Editor can opt to make a decision on a manuscript with two rejects without waiting for additional reviews.



Reviewers are technical experts in the field of the manuscript. TAI aims to provide constructive reviews. TAI expects reviewers to be professional and impartial in their judgement. In some circumstances, a manuscript may get sent to a reviewer who is expert in the wider technical area of a manuscript, without necessarily be an expert in the narrow technical area of the manuscript. This is an example where it is vital for the manuscript to be written in a language to make it accessible to a wider audience.

In addition to providing a written report on each manuscript, reviewers are required to answer a few questions:

  • How would you describe the quality of technical writing used in this manuscript?
  • How would you describe the quality of English used in this manuscript?
  • Are the abstract, introduction, and conclusion written in an accessible language to a non-specialist?
  • Is the work reproducible?
  • How novel is the contribution?
  • How significant is the contribution?
  • Should this manuscript be considered for a 'Best Paper' award?

Some reviewers prefer to load their written report as a file that will be available to authors through Manuscript Central.

Reviews could be resent back to reviewers by the Associate Editor if:

  • The review appears too shallow
  • The comments do not match the decision proposed by the reviewer
  • The review is not objective and/or uses inappropriate language


Decisions on Manuscripts

After reviewers return their reviews with recommendations, the Associate Editor handling the manuscript will write a meta review with a recommendation. The Editor in Chief will make the final decision on the manuscript, which will then be communicated to authors.



Manuscripts with a Minor Revision decision will be allowed 15 days to resubmit their revised manuscript. Manuscripts with a Major Revision decision will be allowed 30 days to resubmit their revised manuscript. Revised manuscripts that are received after these timeframes will need to be submitted as new manuscripts. ScholarOne Manuscripts Central will indicate to authors the number of days left before a deadline passes.


Appeal Process

Authors who wish to appeal a decision, they must submit their appeal in writing to the Editor-in-Chief within 30 days of the decision they appeal. The Editor-in-Chief will make a decision on the appeal within 15 days. Authors should discuss with the Editor-in-Chief if they are still unhappy with the decision on appeal and can escalate the appeal to the IEEE CIS VP Pubs if the discussion with the Editor-in-Chief is unsatisfactory.


Expecteed Time for Review 

We aim to provide speedy decisions on submissions. Associate Editors are requested to handle papers while aiming for a decision within seven weeeks. Our average weeks from submission to first decision sits around 12 weeks mark. However, each submission is unique and may take a different amount of time to review; therefore, outliers exist, unfortunately. The editorial assistant and Editor-in-Chief do their best to follow up with reviewers and Associate Editors.