TL;DR

• Disclose every AI tool you used, always, in every submission.

• AI cannot be listed as an author under any major journal policy.

• Using AI to fabricate data or sources is research misconduct.

• Most journals now have explicit AI policies — read them before submitting.

• Responsible AI use speeds up your work without compromising your integrity.

You have probably used AI to help with your research. Maybe you used it to summarise a dense paper, generate an outline, or clean up a paragraph. That is not automatically wrong. But journals are paying close attention, and the rules around how to use AI in your research without violating journal ethics policies are tightening fast.

The problem is not the tool. The problem is the silence around it. Students submit papers that were partially written or structured by AI and say nothing. Reviewers are increasingly trained to spot this. Editors are adding AI-disclosure requirements. And the consequences of getting it wrong range from rejection to retraction.

This post explains exactly what the current policies say, where the lines are, and how you can use AI responsibly without putting your work or your reputation at risk.

What journal ethics policies actually say about AI

Most major publishers now require authors to disclose any use of AI writing or AI-assisted tools in the preparation of a manuscript. This applies to tools that generate text, paraphrase content, or produce summaries. The Committee on Publication Ethics (COPE), which sets the ethical standards that most peer-reviewed journals follow, states clearly that AI tools cannot be listed as authors because authorship requires accountability, and AI cannot be held accountable for research errors.

COPE published its position on AI and authorship in 2023, and it has been adopted by publishers including Elsevier, Springer Nature, and Wiley. The core requirement across all of them is the same: if you used an AI tool to help write or prepare your manuscript, you must say so in the paper itself, usually in the methods section or an author statement. The disclosure must name the tool and describe how it was used.

What counts as a violation varies slightly by journal. Some journals ban AI-generated text entirely. Others allow it if disclosed. A smaller number have no policy yet, though that gap is closing. Before you submit anywhere, read the journal's author guidelines from start to finish. This is not optional. It is the single most important step you can take to protect your submission.

If you are still working out which journal fits your research, the guide on how to choose the right journal for your research paper walks through the key factors, including how to read author guidelines properly.

Where the line is between acceptable and unacceptable AI use

Acceptable AI use in research means using AI as a tool to support your own thinking, not to replace it. Unacceptable use means presenting AI-generated content as your own original work, using AI to fabricate citations or data, or failing to disclose AI involvement when a journal requires it. The distinction is straightforward, but it is easy to drift across the line without realising it.

Here is a clear breakdown of where most journals draw the boundary:

1. Generally acceptable: Using AI to check grammar and spelling, to summarise papers you have already read and understood, to generate an outline you then rewrite in your own words, or to help format references.

2. Requires disclosure: Using AI to draft any section of the manuscript, to paraphrase your own notes into prose, or to generate explanatory text that appears in the final paper.

3. Not acceptable under any circumstances: Using AI to generate data, fabricate quotations, invent citations, or produce results you present as your own experimental findings. This is research misconduct regardless of whether AI was involved.

The key test is this: can you stand behind every claim, every sentence, and every source in your paper as something you understand, verified, and take responsibility for? If the answer is no for any part of it, that part needs to be rewritten or removed before submission.

If you are preparing your first submission and want to understand the full process, the post on how to submit a research paper to a peer-reviewed journal covers each stage in detail.

How to disclose AI use correctly in your manuscript

Disclosing AI use is not complicated, but it needs to be specific. A vague sentence saying you used AI tools is not enough for most journals. The disclosure should name the tool, state the version if relevant, and describe the specific task it performed.

A correct disclosure looks something like this: in the methods section or a dedicated author statement, write that you used a named AI tool to assist with drafting the literature review section, and that all content was subsequently reviewed, edited, and verified by the authors. That is the level of specificity journals expect.

Here is a simple three-step process for handling disclosure properly:

1. Keep a log as you write. Note every instance where you used an AI tool, what you asked it to do, and what you kept or changed from its output.

2. Check the target journal's author guidelines for their specific disclosure format. Some journals have a dedicated field in their submission system. Others ask for a statement within the manuscript.

3. Write the disclosure in plain language. Do not bury it or minimise it. Reviewers and editors appreciate transparency, and a clear disclosure rarely hurts a submission the way a hidden one can.

Publication Compass is a platform designed to help student researchers navigate exactly this kind of complexity. It helps you identify the right journals for your work and understand their specific requirements before you submit, so you are not reading the fine print at the last minute.

Why AI authorship is never allowed, and what that means for your paper

AI tools cannot be listed as authors on any peer-reviewed paper. This is not a preference. It is a firm policy across all major publishers and is grounded in the definition of authorship itself. Authorship requires that a person made a substantial intellectual contribution, can approve the final version of the manuscript, and can be held accountable for the work if errors are found. AI meets none of these criteria.

This matters for student researchers because the temptation to lean heavily on AI is real, especially when a topic is unfamiliar or a deadline is close. But the more of your paper that is generated by AI rather than authored by you, the less the paper reflects your own research contribution. That is a problem both ethically and practically, because journals are reviewing papers to assess the contribution of the human researchers behind them.

Some students worry that disclosing AI use will make their work look less credible. The opposite is usually true. Transparency signals that you understand research ethics and take them seriously. That is exactly the impression you want to make on a journal editor reviewing your submission.

For high school researchers looking for journals that are well-suited to student work and have clear submission policies, the overview of best peer-reviewed journals for high school researchers is a useful starting point.

How to use AI in your research without violating journal ethics policies in practice

Knowing the rules is one thing. Building good habits into your actual research workflow is another. Here is a practical approach that keeps your work ethical and your submissions clean.

1. Start with your own thinking. Before you open any AI tool, write down what you already know about your topic, what your argument is, and what evidence you have. AI should respond to your ideas, not generate them for you.

2. Use AI for low-stakes tasks first. Formatting, grammar checks, and summarising papers you have already read are the lowest-risk uses. Build your comfort with the tool there before using it anywhere near your core argument or findings.

3. Verify every source independently. AI tools sometimes produce plausible-sounding but incorrect citations. Every reference in your paper must be something you have checked yourself, in the original source. This is non-negotiable.

4. Read your AI-assisted sections aloud. If a paragraph does not sound like you, it probably was not really written by you. Rewrite it until it reflects your own voice and your own understanding of the material.

5. Submit only what you can defend. If a reviewer asked you to explain any sentence in your paper, you should be able to do it. If there are sections where you could not, those sections need more of your own work before the paper is ready.

If you want to understand how specific journals approach student submissions, the detailed guide on the Journal of Student Research scope, requirements, and submission process is worth reading alongside the journal's own guidelines.

How to use AI in your research without violating journal ethics policies when policies are unclear

Some journals, particularly newer or smaller ones, have not yet published a formal AI policy. This does not mean anything goes. When a journal has no stated policy, the safest approach is to follow the COPE guidelines as your default standard, disclose AI use as you would for any journal that requires it, and contact the journal editor directly if you are unsure. Most editors will respond, and asking the question demonstrates exactly the kind of ethical awareness that makes a good impression.

Be especially careful with journals that are not indexed in recognised databases. Some publications have minimal editorial oversight and may not flag ethical issues at submission, but that does not protect you if problems surface later. The post on predatory journals to avoid as a student researcher covers the warning signs in detail.

If you are ready to find a platform that helps you match your paper to the right journal and understand its specific requirements before you submit, joining the waitlist at Publication Compass puts you first in line when the platform opens.

Frequently asked questions

Do I have to disclose AI use if I only used it to fix grammar?

Most major journals require disclosure for any AI tool used in preparing the manuscript, including grammar and language tools. Elsevier and Springer Nature both specify this in their current author guidelines. Check the specific journal's policy, but when in doubt, disclose. A brief, accurate statement costs nothing and protects you.

Can I list an AI tool as a co-author to acknowledge its contribution?

No. AI tools cannot be listed as authors under any current major publisher policy. COPE, Elsevier, Springer Nature, and Wiley all prohibit this explicitly. Authorship requires human accountability. Acknowledge the tool in your methods or author statement instead, and describe what it was used for.

What happens if I submit a paper with undisclosed AI use?

If a journal discovers undisclosed AI use after submission, outcomes can include desk rejection, withdrawal from the review process, or retraction if the paper was already published. The reputational consequences are serious for early-career researchers. Disclosure before submission is always the right choice.

How to use AI in your research without violating journal ethics policies if my school has its own rules?

School and journal policies are separate. You may need to comply with both. If your school prohibits AI use in academic work, that applies regardless of what a journal allows. Read both sets of guidelines and follow the stricter one. If they conflict, ask your teacher or supervisor for written guidance before proceeding.

Are there journals that ban AI use entirely?

Yes. Some journals prohibit AI-generated text in any form, even with disclosure. Science and certain humanities journals have taken this position. Always read the author guidelines of your target journal before writing, not after. Policies are updated regularly, so check the current version each time you submit.

The bottom line

Using AI in your research is not inherently wrong. Using it without understanding the rules, without disclosing it, or without taking responsibility for the content it produces is where researchers get into trouble. The path forward is straightforward: know the policy of every journal you submit to, disclose AI use clearly and specifically, and make sure every word in your paper is something you can stand behind.

The researchers who build strong publication records are not the ones who avoid AI entirely. They are the ones who use it carefully, honestly, and within the boundaries that keep academic work trustworthy. For more on navigating the full research and publication process, visit the Publication Compass blog.