AI journals have multiplied alongside the field itself, and for researchers that is both good news and a real problem. There are more places than ever to publish artificial intelligence research, but the scope, quality and credibility of these journals vary enormously. Choosing the wrong one can cost you months of review time, a rejection on scope grounds, or a paper that the right readers never find.
This guide explains what AI journals are, how they differ from conferences, and the practical criteria for comparing them. It also shows why the International Journal of Advanced Computer Science and Applications (IJACSA), an established open access journal for computer science and AI research publishing since 2010, is worth a place on your shortlist, and how to submit if it matches your work.
Disclosure: this article is published by The Science and Information (SAI) Organization, which also publishes IJACSA.
Key Takeaways
The strongest AI journals combine a clear scope, recognised indexing, transparent peer review, and open access with persistent identifiers such as DOIs.
Compare AI journals on scope, indexing, peer review, access and cost, and transparency, in that order.
Indexing is selective: Elsevier says only about a third of titles suggested to Scopus pass technical screening, and only half of those are accepted (Elsevier, "Scopus content," July 2025). Verify any claim in Scopus or Web of Science yourself.
AI publishing is growing fast: AI publications rose from about 102,000 in 2013 to more than 242,000 in 2023 (Stanford HAI, "AI Index Report 2025"), so the venue you pick is a strong first signal to readers.
IJACSA has published peer-reviewed computer science and AI research every month since 2010, as a gold open access journal under CC BY 4.0, with DOIs registered through Crossref.
What Are AI Journals, Exactly?
AI journals are peer-reviewed publications that publish research on artificial intelligence and its subfields, such as machine learning, natural language processing, computer vision, knowledge representation and robotics. Some focus on a single subfield, while others cover AI alongside the rest of computer science.
Either way, the best ones share a few traits: a clear scope, a transparent review process, recognised indexing and a long, verifiable publishing record. That last point deserves attention. A journal that has published on a regular schedule for many years has a track record you can inspect: its archive, its editorial board, its citation data and the researchers who have chosen to publish there.
The right choice still depends on what your paper contributes and who needs to read it, so the sections below show how to compare AI journals against the same criteria.
Why the Choice of AI Journals Matters More Than Ever
The volume of AI research is growing quickly. The Stanford AI Index reports that AI publications grew from approximately 102,000 in 2013 to more than 242,000 in 2023, and that AI's share of all computer science publications rose from 21.6% to 41.8% over the same period (Stanford HAI, "AI Index Report 2025," Research and Development chapter).
For authors, that growth has two consequences. First, readers have far more to sift through, so the venue you choose is often the first signal of whether your paper deserves their attention. Second, reviewers and editors at the most competitive journals are stretched, which makes a poorly matched submission more likely to be rejected quickly. Spending an hour on journal selection before you submit usually saves weeks afterwards.
Journals vs. Conferences: Where AI Research Gets Published
AI is unusual among research fields because conferences and preprints carry so much weight. arXiv recorded a then-record 24,226 new submissions in October 2024, with computer science the largest category (arXiv blog, "arXiv sets new record for monthly submissions again," November 2024). Major venues are also open: the NeurIPS proceedings archive is freely readable, and the Journal of Machine Learning Research makes all its papers freely available.
A practical way to decide: conferences and preprints suit fast-moving results that benefit from quick feedback, while journals suit complete studies, surveys, reviews and work that needs space for full methods and evaluation. Many universities and funders also look at where a paper is published, so a journal indexed in Scopus or Web of Science is often the safer choice for promotion, degree requirements and funding applications. Check your own institution's rules before you decide.
Five Criteria for Comparing AI Journals
Whichever shortlist of AI journals you build, run each one through the same five checks.
Scope and fit. Read the aims and scope, then skim the last two or three issues. If no recent paper resembles yours in topic and type (original research, survey, application), look elsewhere.
Indexing and visibility. Check whether the journal is indexed in services your institution recognises, such as Scopus or Web of Science, and whether each article receives a DOI registered through Crossref so that it can be found and cited reliably. Our guide to verifying a Scopus journal explains how.
Peer review. Look for a clearly described process: how many reviewers, whether review is single-blind or double-blind, and what the journal evaluates (originality, correctness, relevance, readability).
Access and cost. Find out whether the journal is open access, which licence applies, and whether there is an article processing charge (APC). Reputable journals state fees openly before you submit.
Transparency. You should be able to identify the publisher, contact the editorial office, and see a named editorial board with real affiliations.
How to Spot Trustworthy AI Journals Before You Submit
Before you submit, run a quick credibility check. The Think. Check. Submit. initiative gives authors a checklist for this. Its questions are simple: Do you or your colleagues know the journal? Can you easily identify and contact the publisher? Is the peer-review process clearly described? Are the editorial board members identifiable and genuinely affiliated? Is the journal indexed where it claims to be, and is it transparent about fees?
Warning signs to watch for include guaranteed fast acceptance, unsolicited invitations that flatter your work, journal names that closely imitate well-known titles, and fees that only appear after acceptance. Check indexing directly in Scopus or the Web of Science Master Journal List rather than relying on a logo on the journal homepage.
Selectivity is a useful benchmark here. Elsevier states that every year thousands of new titles are suggested for Scopus, but only about 33% meet its technical criteria, and of the roughly 1,200 that do, only about 50% are accepted after review by its independent Content Selection and Advisory Board (Elsevier, "Scopus content," July 2025). A badge on a homepage tells you none of that.
Open Access vs. Subscription AI Journals: What It Means for Your Paper
Open access AI journals make every article free to read from the day it is published. That helps your work reach students, practitioners and researchers whose institutions lack a subscription, and many funders now require it. Gold open access made up 44% of global articles, reviews and conference papers in 2025, up from far lower levels a decade earlier (STM Association, "Open Access Dashboard: Uptake of Open Access," 2025).
The usual trade-off is that open access journals cover their costs through APCs paid by authors or their funders, while subscription journals typically charge authors little or nothing but place articles behind a paywall. Before choosing, check your funder and institution policies, and look at the licence. Gold open access combined with a Creative Commons Attribution 4.0 International License (CC BY 4.0) lets others reuse your work with attribution, which is often what authors want. Our guide to open access journals covers the models and costs in more detail.
Why IJACSA Stands Out Among AI Journals: Established Since 2010
The International Journal of Advanced Computer Science and Applications is a computer science journal that publishes AI research and has been doing so since 2010, so its track record is long enough to inspect for yourself. Artificial intelligence and machine learning sit at the front of its scope, alongside data science, computational intelligence and network security, which makes it a natural home for AI research with a computing focus.
Over more than fifteen years of monthly publishing, it has built a reputation you can verify rather than take on trust: dual indexing in Scopus and Web of Science, a public editorial board drawn from universities around the world, a complete archive of every issue, and a community of authors and reviewers across the globe.
Against the five criteria above, IJACSA looks like this:
Indexing: indexed in Scopus (CiteScore 3.4, Q3) and in the Web of Science Emerging Sources Citation Index, along with EI Compendex, Inspec and other databases.
Peer review: double-blind, with each article reviewed by at least three independent reviewers. The acceptance rate is approximately 15%.
Access: a gold open access journal, with every article free to read and download immediately on publication under CC BY 4.0. Authors keep the copyright to their work. An APC applies to accepted papers, and the current fee schedule is published on the Call for Papers page.
DOIs and Crossref: SAI is a Crossref member, and every published article is assigned a DOI registered through Crossref, which gives each paper a permanent link and makes citations easier to track.
Transparency: the editor-in-chief, full editorial board and editorial office contact are listed publicly, and every past issue is available online.
Reading Published Papers in AI Journals: Start With the Archives
Not every visitor is looking to submit. If you want to read what has already been published, for a literature review, a benchmark comparison or simply to see what current AI research looks like, the fastest route is an archive that lets you browse by issue and read full text without a login.
The IJACSA archive. Every volume and issue published since 2010 is listed by year, and each issue opens to its table of contents. AI and machine learning papers appear alongside other computer science topics, so scan the titles in each issue for what matches your interest.
The IJARAI archive. The International Journal of Advanced Research in Artificial Intelligence was SAI's dedicated AI title. It no longer accepts submissions, but its full-text archive remains free to read, which makes it useful if your interest is AI specifically.
Each article page gives the full text, the authors and a DOI you can use to cite it, and every article is open access under CC BY 4.0, so you can read, download and reuse it with attribution.
Types of Papers Accepted by AI Journals
AI journals do not all take the same kinds of manuscript, and the article type shapes both the structure of your paper and what reviewers look for. Knowing the common types helps you match your work to the right venue before you write the final draft.
Original research. A new method, model, dataset or theoretical result, judged on novelty, correctness and evidence. Expect to be asked for comparisons against current baselines.
Survey and review papers. A structured overview of a subfield that organises existing work, compares approaches and points out open problems. These are judged on coverage, accuracy and insight rather than new experiments, and they are especially useful in fast-growing areas.
Application papers. Work that applies established techniques to a real domain such as healthcare, finance, education or engineering, judged on how well the evaluation reflects real conditions.
Replication and negative-result papers. Studies that test whether earlier claims hold up. Their value is rising as reproducibility becomes a bigger concern.
Dataset and benchmark papers. Contributions that give other researchers new data or a fairer way to compare methods.
Journals differ in which of these they welcome, so check the scope statement and recent issues rather than assuming. Choosing a journal that regularly publishes your article type is one of the simplest ways to avoid a rejection on scope grounds.
Preparing Your Manuscript for AI Journals
Good preparation improves your chances at any journal you target, and most of it is not specific to one venue.
State the contribution in the abstract. Say what is new, how you tested it, and what the results show, in plain terms.
Make results reproducible. Describe datasets, preprocessing, model settings and evaluation metrics, and compare against current baselines rather than only older ones. A 2025 review in AI Magazine maps the barriers to reproducibility across methods, code, data and experiments, and finds that reproductions often differ substantially from the originals (Semmelrock et al., "Reproducibility in machine-learning-based research," AI Magazine, 2025).
Be open about limitations and ethics. Reviewers expect a candid discussion of where the method fails, along with any data-privacy or bias considerations.
Follow the formatting rules. Read the author guidelines before you write the final version, not after.
Anonymise for double-blind review. Remove author names, affiliations and identifying self-references from the manuscript file.
Submit to one journal at a time. Most journals, this one included, do not consider manuscripts that are under review elsewhere.
For IJACSA specifically, read the author guidelines, check the schedule on the Call for Papers page, and submit your manuscript when it is ready.
Common Mistakes When Submitting to AI Journals
Ignoring scope. A strong paper sent to the wrong journal is still rejected. Match your topic to what the journal actually publishes.
Chasing impact numbers alone. A single metric says little about fit. Consider readership, indexing and review quality together.
Overlooking fees and licences. Read the APC and copyright terms before you submit, not after acceptance.
Weak evaluation. Papers with a single dataset, no baselines or no statistical care struggle at almost every serious venue.
Submitting too early. Ask a colleague outside your project to read the draft first. Unclear writing is a common reason for avoidable rejections.
Frequently Asked Questions About AI Journals
How do I choose between AI journals?
Start with scope: pick journals whose recent papers resemble yours. Then compare indexing, peer-review process, access model and fees, and verify each claim directly in the indexing database or on the publisher's site.
Should I publish AI research in a journal or at a conference?
Conferences suit fast-moving results that benefit from quick feedback. Journals suit complete studies, surveys and reviews that need room for full methods and evaluation, and a journal indexed in Scopus or Web of Science is often what institutions and funders look for. Check your own requirements first.
Is IJACSA an AI journal, and is it indexed in Scopus and Web of Science?
It is a computer science journal that publishes AI and machine learning research as a core part of its scope, so it works as a home for AI papers with a computing focus. It is indexed in Scopus (CiteScore 3.4, Q3) and in the Web of Science Emerging Sources Citation Index.
Do open access AI journals charge to publish?
Many gold open access journals charge an article processing charge, though some do not. IJACSA's articles are free for everyone to read, and an APC applies to accepted papers, with the current schedule published in the Publishing Fees section of its Call for Papers page.
Does IJACSA assign a DOI to published articles?
Yes. Every article is assigned a DOI registered through Crossref, giving it a permanent link that is easy to cite and track.
Can I verify a journal's Scopus status myself?
Yes. Search the exact title or ISSN on the official Scopus source page and confirm the listing is current, not historical. Never rely on a badge on the journal's own site.
Next Steps: Choosing Among AI Journals
Choosing among AI journals is less about finding the single best title and more about finding the one that fits your paper, your readers, and your requirements for indexing, access and cost. Build a short list, compare each journal against the same five criteria, and verify every claim yourself.
If your work falls within computer science and artificial intelligence, review the IJACSA Call for Papers for scope, dates, fees and how to submit. If you came to read rather than write, browse the archive above; the full text of every paper is free.
