Conflict of interest (COI) disclosure is the process by which authors, and often reviewers and editors, report relationships or interests that could plausibly bias — or be perceived to bias — the research, its interpretation, or its publication. It sits alongside, but is distinct from, contribution disclosure mechanisms like CRediT: CRediT records what someone did on a paper; COI disclosure records what relationships might have influenced how they did it, or how the work was judged. Journals require both, for related but separate reasons, and getting either one wrong at submission is one of the more common — and most avoidable — reasons a paper stalls in editorial handling.
This guide covers what actually counts as a conflict of interest in scholarly publication, what the field’s standard reference document (the ICMJE Disclosure Form) requires, when and how to disclose, what happens when disclosure is skipped, and how COI disclosure connects to CRediT-style contribution transparency.
Financial vs. non-financial conflicts of interest — what counts
The International Committee of Medical Journal Editors (ICMJE) — whose recommendations are followed far more widely than the small group of journals formally affiliated with it — defines a conflict of interest as existing “when professional judgment concerning a primary interest (such as patients’ welfare, or the validity of research) may be influenced by a secondary interest.” Critically, ICMJE is explicit that perceptions of a conflict are as important as actual conflicts — a relationship doesn’t need to have actually changed anyone’s behavior to be disclosable; it only needs to be the kind of relationship a reasonable reader might think could have.
That definition splits into two categories, and journals generally ask about both:
Financial relationships
- Employment (including a co-author’s employer, if it’s a company with a stake in the outcome)
- Consultancies and advisory-board roles
- Stock ownership or stock options
- Honoraria for speaking, teaching, or manuscript preparation
- Patents (issued, pending, or licensed) related to the work
- Paid expert testimony
- Research funding and grants from an entity with a stake in the results, including in-kind support (e.g., a manufacturer supplying study drug or equipment)
Non-financial relationships
- Personal relationships or rivalries with other authors, reviewers, or subjects of the research
- Academic competition (e.g., a competing lab racing toward the same finding)
- Institutional or advocacy affiliations that could shape interpretation (board membership on an advocacy organization, a role in a patient-advocacy group with a stake in a particular outcome)
- Strongly held intellectual beliefs or prior public positions that could make objective evaluation of a study’s findings difficult
A useful test: would disclosure of this relationship change how a reasonable reader interpreted the paper’s conclusions, even slightly? If yes, disclose it — the cost of an unnecessary disclosure is nearly zero; the cost of an omitted one, described below, is not.
What journals typically require: the ICMJE Disclosure Form
Most biomedical and many non-biomedical journals ask authors to complete the ICMJE Uniform Disclosure Form for Potential Conflicts of Interest (or a journal-specific form modeled directly on it) at submission. ICMJE originally called this a “conflicts of interest” form; it was renamed to the “disclosure form” in February 2021 (formalized June 2021) — partly to reduce the negative connotation of the word “conflict” for relationships that are common, disclosable, and not inherently disqualifying, and partly to broaden the range of relationships and activities the form asks about.
The form is organized around identifying information, the specific work under consideration, the author’s financial relationships and activities outside that submitted work, and other relationships or activities a reader could reasonably perceive as having influenced it. Each author on a manuscript typically completes a separate copy — disclosure is an individual, not a corresponding-author-only, obligation. One procedural detail worth knowing: ICMJE is explicit that a completed form is not submitted to ICMJE itself and cannot be saved on its website — authors download it, complete it, and submit it directly to the journal per that journal’s own instructions.
Because ICMJE’s recommendations are a reference standard rather than a binding mandate, adoption varies: some journals require the ICMJE form verbatim, some require an equivalent journal-specific COI statement, and some ask for a shorter disclosure paragraph within the manuscript itself rather than a separate form. Authors submitting to a journal for the first time should check that journal’s own author instructions rather than assume the ICMJE form applies by default — and should be cautious of journals that claim to follow ICMJE recommendations without actually using a current, ICMJE-consistent disclosure process.
When and how to disclose
- At submission. COI disclosure is expected as part of the initial manuscript submission, not something added later if an editor asks. Most journal submission systems require a completed disclosure form or statement before the manuscript can proceed to review.
- Kept current through revision. If a relationship changes between initial submission and acceptance — a new consultancy begins, a patent is filed — ICMJE’s guidance is that this needs to be reflected, not left as it stood at first submission.
- In the published article. Disclosed relationships appear in the published version, typically in a dedicated “Conflicts of Interest” or “Competing Interests” section, distinct from acknowledgments and from any CRediT contribution statement.
- Reviewers and editors disclose too. ICMJE’s recommendations extend the same obligation to peer reviewers, who must disclose to editors any relationship that could bias their opinion of a manuscript and recuse themselves where a real conflict exists, and to editors handling a manuscript, who should recuse themselves from the final decision where they have a relevant conflict. Disclosure in scholarly publication is not solely an author-facing requirement.
What happens when a conflict isn’t disclosed
ICMJE states this plainly: “Purposeful failure to report those relationships or activities specified on the journal’s disclosure form is a form of misconduct.” That framing — misconduct, not merely an administrative omission — is why undisclosed COI is treated seriously once discovered, whether that discovery happens during peer review, after publication via a reader or a member of the community, or during an institutional investigation triggered by something else entirely.
The Committee on Publication Ethics (COPE) provides case guidance journals use to work through exactly this scenario, and its current retraction guidance is explicit on the consequence: COPE’s Retraction Guidelines (Version 3, 2025) name undisclosed major conflict of interest as one of the recognized grounds for retracting a published article — alongside fabrication, falsification, and plagiarism. In practice this means an omitted disclosure isn’t necessarily fixed with a quiet correction; depending on how material the undisclosed relationship is judged to be, the outcome can range from a published correction or an added disclosure statement, up to a formal retraction (see CASRAI’s guide to how a retraction actually happens for the editorial process that follows). Institutions may separately treat an undisclosed significant financial interest as a research-integrity or financial-conflict-of-interest compliance matter under their own policies, independent of whatever action the journal takes.
How this connects to CRediT-style contribution transparency
COI disclosure and CRediT contribution statements are often completed at the same point in a submission workflow and can look, to an author filling out forms quickly, like variations on the same task. They aren’t. CRediT (the Contributor Roles Taxonomy, standardized as ANSI/NISO Z39.104-2022, which CASRAI originated and now co-stewards with NISO) records what each named author actually contributed — conceptualization, data curation, formal analysis, funding acquisition, and the taxonomy’s other defined roles. COI disclosure records what outside relationships might have shaped that contribution.
The two are complementary rather than redundant, and reading them together often tells a reader more than either alone: a CRediT statement showing a pharmaceutical-company employee held the “Formal Analysis” and “Writing – Original Draft” roles on a trial evaluating that company’s own product is a different signal, read alongside that same author’s disclosed employment and stock-option relationship in the COI statement, than either fact would be on its own. This is also why journals that have adopted CRediT increasingly present both disclosures in the same transparency block near the end of an article, rather than burying one in an acknowledgments paragraph and the other in a supplementary file. For a closer look at how CRediT statements themselves are structured and worded, see CASRAI’s CRediT author statement samples guide.
Frequently asked questions
Is having a conflict of interest the same as having done something wrong?
No. Financial and non-financial relationships covered by COI disclosure — employment, past consulting work, grant funding — are extremely common in active research careers and are not, by themselves, evidence of misconduct or of a biased result. What ICMJE and COPE treat as misconduct is the failure to disclose a relevant relationship, not the relationship’s existence.
Do all co-authors need to disclose, or only the corresponding author?
Each author is generally expected to complete their own disclosure. A corresponding author submitting a single combined statement without confirming each co-author’s individual relationships risks omitting something a co-author should have reported.
What if I genuinely have nothing to disclose?
State that explicitly (“The authors declare no competing interests” or the journal’s equivalent standard phrasing) rather than leaving the section blank — an affirmative statement of no conflict is itself part of the disclosure record, not just a null response.
Does funding from a government agency or non-profit need to be disclosed?
Yes — disclosure forms typically ask about all funding sources for the specific work, not only ones with a plausible bias. The distinction that matters for whether a relationship rises to a “conflict” (as opposed to simply being disclosed as a funding source) is usually whether the funder has a stake in a particular result, which a public funder or non-profit typically does not, but the funding relationship itself still gets reported.
Is the ICMJE form legally binding?
No. ICMJE is a voluntary group of journal editors, and its recommendations are guidance that journals choose to adopt, not law or regulation. That said, a journal that has adopted ICMJE’s disclosure requirements as a condition of publication can reject, correct, or retract work over noncompliance, and an institution can separately treat non-disclosure as a violation of its own research-integrity or financial-conflict-of-interest policy regardless of what any journal does.
Related CASRAI resources
- Research integrity & compliance — the cluster hub for this and related compliance topics
- Conflict of interest disclosure — the CASRAI Dictionary’s operational definition, examples, and counter-examples
- COPE (Committee on Publication Ethics)
- Retraction and how a retraction actually happens
- Research misconduct
- Clinical trial registration and reporting compliance
- CRediT & authorship and CRediT author statement samples
