IEEE Plagiarism Policy and Prohibited Authors List (PAL)

IEEE (the Institute of Electrical and Electronics Engineers) is one of the few major scholarly publishers to publish a fully tiered, numerically-scaled plagiarism severity framework rather than leaving investigators to weigh each case from first principles. The framework — documented in IEEE’s Publication Services and Products Board (PSPB) Operations Manual and summarized in IEEE’s own ‘Plagiarism Levels and Corrective Actions’ reference — sorts a substantiated plagiarism finding into one of five levels based on how much text was copied, whether it was credited at all, and whether it was properly delineated as a quotation. The practical consequence for authors sits at the other end of the same policy: a suspension from submitting to IEEE escalates, for the duration of that suspension, into inclusion on IEEE’s internal author-screening list — widely known as the Prohibited Authors List (PAL), and referred to in more recent PSPB Operations Manual language as the PPL.

The five plagiarism levels

IEEE grades a substantiated case using the following criteria, from most to least severe:

• Level One — uncredited verbatim copying of an entire paper, or of a major portion (more than 50%) of a single paper, or verbatim copying that recurs across more than one paper by the same author(s).
• Level Two — uncredited verbatim copying of a large portion of a paper, defined as roughly 20-50% of the text, or recurring verbatim copying across more than one paper at that scale.
• Level Three — uncredited verbatim copying of individual elements (a paragraph, a sentence, a figure or illustration) that together add up to a significant but smaller portion of the paper, generally up to about 20%.
• Level Four — uncredited but non-verbatim reuse: improper paraphrasing of pages or paragraphs from another source, close enough to the original that it does not represent the author’s own analysis or wording, without attribution.
• Level Five — verbatim copying of a major portion of a source that IS credited (the original is cited somewhere in the paper) but is not clearly set off as a direct quotation — no quotation marks, block-indent, or other marking that would tell a reader exactly which words are copied text rather than the author’s own.

The scale is deliberately about extent and disclosure, not intent — IEEE’s investigation process determines the level from what was copied and how it was (or wasn’t) marked, independent of whether the author claims the overlap was accidental. This is consistent with how similarity-checking tools like Turnitin or iThenticate are used across scholarly publishing generally: the raw percentage a checker returns is a starting signal for a human reviewer, not a self-executing verdict, but IEEE is unusual in having published an explicit, numbered scale that translates a substantiated finding directly into a specific tier of consequence.

Corrective actions and self-plagiarism

Consequences scale with the level and with an author’s prior history. Public IEEE Communications Society plagiarism-policy guidance, for example, describes actions ranging from a reminder that the practice is inappropriate scholarly conduct, through a required letter of apology to the venue’s editor, up to a suspension of submission privileges to a specific IEEE publication — commonly cited minimums run around six months to one year for a first, lower-level finding, with cited ranges up to five years for more serious cases. Repeated Level One findings by the same author increase the suspension duration further, and the PSPB Chair has authority to administer a lifetime prohibition on publishing in any IEEE-copyrighted venue for authors with a pattern of serious, repeated violations.

IEEE treats undisclosed reuse of an author’s own prior text — self-plagiarism or text recycling — as a related but analytically separate concern from the Level One-Five scale described above, which is built around copying another party’s material. Self-plagiarism findings are addressed under IEEE’s own text-recycling guidance and can still trigger corrective action, but they are not scored on the same uncredited-copying levels; a properly cited reuse of an author’s own earlier methods description, for instance, is a text-recycling judgment call for the editor, not an automatic plagiarism-levels case. See CASRAI’s self-plagiarism entry for how the distinction between self-plagiarism and third-party plagiarism plays out more generally, including why the US Office of Research Integrity’s federal misconduct definition does not treat self-plagiarism as research misconduct even where a publisher’s ethics policy still treats it as a publication-ethics violation.

The Prohibited Authors List (PAL / PPL)

When an author is placed under a submission suspension, IEEE adds that author’s name and known email addresses to an internal screening database. This list has long been referred to across IEEE’s own published guidance — including PDFs titled ‘Appropriate Use of the IEEE Prohibited Authors List’ and a PAL Database FAQ sometimes informally called the ‘Banned Author List’ — as the Prohibited Authors List (PAL). More recent PSPB Operations Manual language has formally relabeled it the PPL, though ‘PAL’ remains the term most publication volunteers and the wider community still use, and IEEE’s own reference documents continue to carry the PAL name.

In practice, the list functions as an editorial screening tool rather than a public register: IEEE publication volunteers — journal editors, conference program chairs, and their editorial staff — can check a submitting author’s name and email address against the PAL/PPL database before a manuscript proceeds to review, either by checking an individual name or uploading a batch of submitted authors for comparison. A name match is a prompt to escalate, not a final determination: IEEE’s own guidance directs volunteers who find a matching name to contact IEEE’s Intellectual Property Rights (IPR) Office to confirm the author’s identity before declining or withdrawing a submission on that basis, since the database is matched on name and email rather than a verified unique identifier. This is one of the practical reasons some IEEE societies and journals increasingly reference persistent author identifiers such as ORCID alongside submission metadata — a stable identifier reduces the ambiguity a name/email match alone can introduce.

A PAL/PPL listing is time-bound and tied to the specific suspension that produced it: it applies for the duration of the sanction determined by the level of the finding (and any escalation for repeat Level One offenses), not indefinitely, except in the rare lifetime-prohibition cases the PSPB Chair can impose for the most serious repeat violations.

Why this matters for research administrators and authors

For research offices, authorship-integrity officers, and co-authors of a submission under investigation, the practical stakes of IEEE’s framework are twofold. First, because IEEE grades severity by extent and disclosure rather than a single fixed threshold, a similarity-checker percentage alone does not predict the outcome — the same raw percentage can land at a different level (or no violation at all) depending on whether the overlap is properly quoted, concentrated in one section versus spread across the paper, or recurs across multiple submissions by the same author. Second, because a PAL/PPL suspension is venue- and time-bound rather than a single IEEE-wide flag, co-authors on a paper by someone under suspension should confirm directly with the relevant IEEE society or journal whether a specific submission is affected, rather than assuming a suspension at one IEEE venue automatically extends to every other IEEE-copyrighted publication.

IEEE’s editorial ethics policies — of which the plagiarism-levels scale and PAL/PPL are one part — are set partly IEEE-wide and partly at the level of the individual technical society, journal, or conference, the same pattern CASRAI’s IEEE Double-Blind Peer Review entry describes for peer-review models and its IEEE Generative AI Policy entry describes for AI-disclosure rules: authors and reviewers should always confirm the exact policy text published by the specific venue they are dealing with, rather than assuming uniform IEEE-wide wording.