Information Taken Directly From an Existing Classified Source and Stated Verbatim in a New or Different Document Is an Example of Plagiarism and Classification Marking Failure: What Actually Happens

Taking text verbatim from a classified source and dropping it into a new or different document is an example of two overlapping violations: plagiarism in the broad sense of uncredited copying, and, more consequentially, a derivative classification error — the new document now carries classified information without the markings, access controls, and distribution limits the original carried. The verbatim nature is what makes it a distinct problem. Paraphrasing can strip a source identifier; copying cannot.

That distinction matters because the classification system does not care about intent, and it rarely cares about attribution. It cares about whether the information is protected, whether the new container correctly declares that protection, and whether everyone who touches the new document has the clearance and need-to-know to hold it. A sentence lifted word-for-word from a classified briefing into an unclassified training slide fails all three tests at once.

What follows is the working mechanics: where the rule lives, how derivative classification works, what "verbatim" adds to the offense, what happens in practice when it's caught, and where the common assumptions about it break down.

Where the Rule Actually Lives in US Policy

The controlling document for US executive branch classified information is Executive Order 13526, "Classified National Security Information," signed December 29, 2009. Section 1.4 covers classification, and Section 1.5 governs derivative classification specifically. The relevant sentence: derivative classifiers must "apply the classification markings of the source information" and may only use the source's classification level, duration, and reason.

That means the person copying doesn't get to make an independent judgment. If the source paragraph is Secret with a 25-year duration and the reason "1.4(c)," the new document carries Secret, 25 years, and 1.4(c). No downgrading to Confidential because the new audience seems friendlier. No "UNCLASS" header because the slide deck is going to a contractor.

The Information Security Oversight Office (ISOO) publishes implementing guidance, including the 32 CFR Part 2001 regulation that spells out marking requirements. ISOO's marking handbook — the Standard Form 703-style format familiar to anyone who has handled classified material — specifies that derivative documents must show source of classification, declassification date or event, and the reason, in addition to the level itself.

Compare with the UK. The UK's Government Security Classifications policy, in force since April 2014, uses OFFICIAL, SECRET, and TOP SECRET, and it explicitly requires that any extract or copy retain the classification of the original. The UK's approach is less prescriptive on marking format than the US system but stricter on handling — SECRET material is limited to those with a security clearance and a defined business need.

Derivative Classification: The Mechanism That Makes Verbatim Copying Dangerous

Derivative classification is the act of incorporating classified information into a new document, whether by quoting, paraphrasing, restating, or generating in new form. The US system treats the act as requiring the same care as original classification, and the person doing it becomes a derivative classifier — a role that carries legal obligations under 18 U.S.C. § 798 and, for certain categories, under the Atomic Energy Act.

Three outputs are mandatory for any derivative document: the classification level, the declassification instruction, and the classification authority block identifying who classified it and on what basis. The authority block format is "Derived from: [source], Declassify on: [date or event], Reason: [1.4 letter]." If the new document says "Derived from: multiple sources" without listing them, that's a marking deficiency, not a pass.

Verbatim copying adds a fourth output that people often skip: the source citation in the body itself. The rule doesn't strictly require it, but ISOO's marking guidance recommends that portions carry the source document's classification authority so a future reviewer can trace the lineage. Without that lineage, a later declassification review can't determine whether the copied paragraph is still properly classified — which is exactly why verbatim copies with broken provenance get flagged in inspections.

A worked example. An analyst pulls two sentences from a Secret//NOFORN assessment dated March 2023, with declassification set for 2048, and pastes them into a draft unclassified white paper. Correct handling: the white paper becomes Secret//NOFORN, the authority block cites the assessment, the declassification date is 2048, and the two sentences are marked with a portion marking (S//NF). Incorrect handling — the common one — produces a document that should be Secret but is marked unclassified, which is an unauthorized disclosure waiting to be forwarded.

What Verbatim Copying Adds Beyond Ordinary Plagiarism

Plagiarism is passing off someone else's work as your own. In academic and journalistic contexts it's an ethics and integrity failure, and the remedy is retraction, correction, or discipline. Classified information is a different framework entirely, and the overlap is narrower than people assume.

Two cases share the verbatim element. Case one: copying from a classified source into a new classified document without source attribution. That's a classification marking deficiency, and depending on the source's access controls, it may also be a need-to-know violation. Case two: copying from a classified source into an unclassified document. That's an unauthorized disclosure, period, regardless of whether the copier had clearance.

The second case is the one that catches people. Consider a cleared contractor who writes a training module for a broader audience and lifts one illustrative sentence from a classified contract document because it "wasn't sensitive." The sentence is now in an unclassified product distributed to people without clearance. The classification status of the source governs, not the copier's impression of sensitivity. The information was classified, the new container is not, and the disclosure has already occurred the moment the document leaves the copier's control.

Attribution makes no difference here. A footnote saying "Source: classified report XYZ" would actually make things worse — it flags to the reader that the sentence came from classified material, which itself may be classified metadata. There is no version of "I cited it, so it's fine" in this framework.

The Marking Chain: How a Verbatim Sentence Gets Traced Back

Marking is the paper trail, and the paper trail is what makes verbatim copying traceable. Every properly marked classified document carries a classification authority block, portion markings, and — for US material — a banner line on every page. A verbatim sentence that appears in a later document often retains its original phrasing in a way a paraphrase would not, and that phrasing is searchable.

Security investigators use three techniques. First, textual comparison: automated tools flag long strings of matching text between classified holdings and unclassified documents, similar to plagiarism detection software but applied to classified repositories. Second, source lineage: an inspector compares authority blocks across documents to see whether the same paragraph appears under mismatched classification levels. Third, human recognition: the original author or a colleague notices their sentence in the wrong place.

The traceability matters because it changes the consequence. A paraphrased leak requires the investigator to prove the underlying information is classified. A verbatim leak proves itself — the sentence is identical to the source, the source carries a classification marking, and the chain from classified container to unclassified container is direct. The burden of proof shifts.

Separately, copyright applies to classified material too. US government works are generally not subject to copyright domestically under 17 U.S.C. § 105, but contractor-authored classified reports usually are, and copying verbatim into a derivative work can create a copyright issue layered on top of the security issue. The security issue always dominates.

Consequences in Practice: What Happens When It's Caught

The realistic outcomes fall into four buckets, and the bucket depends on whether the disclosure stayed inside the cleared community or left it.

The classification level matters as much as the destination. A Confidential sentence copied into an unclassified memo is a serious incident but usually handled administratively. A TOP SECRET//SCI sentence in the same position is a different animal — the SCI compartment alone may make the verbatim text a "special access" violation, and the damage assessment will involve the originating program office.

In the US, 18 U.S.C. § 798 criminalizes the disclosure of classified cryptographic and communications intelligence information, with penalties up to 10 years per offense. The Espionage Act (18 U.S.C. § 793) covers unauthorized retention and transmission of national defense information. Most verbatim-copying incidents never reach prosecution — they're resolved through the agency's security process — but the criminal statutes set the ceiling, and investigators sometimes use them as leverage in interviews.

In the UK, the Official Secrets Act 1989 creates criminal offenses for disclosure by Crown servants and contractors, with penalties up to two years for most categories and up to 14 years for certain defense-related disclosures. The UK's approach is narrower in scope but criminalizes the act of disclosure itself rather than requiring a showing of harm in every case.

Why "It Was Already Classified, So It's Fine" Fails

The most common rationalization for verbatim copying from a classified source into a new classified document is that the information was already protected, so nothing new was exposed. That logic breaks in three places.

First, the new document has a different distribution list. The original may have been NOFORN (no foreign nationals), or limited to a specific program, or marked with a special access requirement the new audience doesn't hold. Copying it into a wider classified document expands the audience without a need-to-know determination. The classification level is the same; the access control is not.

Second, the derivative document now has a lifecycle the original didn't. It gets stored, transmitted, archived, and eventually declassified on its own schedule. If the authority block is missing, the declassification review has no basis for evaluating it, and it may sit in the archive longer than the original — or be released prematurely. Both are failures.

Third, the marking requirement is not a suggestion. A correctly classified document with a wrong or missing authority block is a deficiency that will be found in inspection. ISOO publishes self-inspection results annually, and marking deficiencies are consistently among the top findings — not because people don't know the rule, but because derivative copying inherits the source's markings only if the copier transfers them. Verbatim copying makes the omission invisible: the text looks fine, so nobody checks the block.

Self-Classification and the Limits of the Copier's Judgement

People who copy verbatim often believe their own judgment about sensitivity should govern the new document's marking. It doesn't, and the policy is explicit about why. Derivative classifiers apply the source's markings; they don't re-derive them. The authority to classify originally rests with specific officials designated in writing under E.O. 13526 Section 1.3. Everyone else operating downstream is applying, not deciding.

That hierarchy creates a specific failure mode: a cleared employee copies a sentence from a document marked Secret because it "reads like unclassified analysis," places it in an unclassified product, and is later surprised when the source document's classification is confirmed. The employee's read of the sentence's sensitivity is irrelevant. What governs is the source's marking, which was set by someone with original classification authority.

Self-classification also cuts the other way. Someone who copies from an unclassified source and then marks the new document Secret "to be safe" has created a different problem — over-classification, which ISOO tracks as its own deficiency category and which wastes review resources and delays legitimate release. The rule is symmetrical: apply the source's marking, not your impression of it.

Handling a Verbatim Copy You've Already Made

If you realize after the fact that you've pasted classified text into the wrong container, the sequence matters. The wrong move is to quietly edit the document and hope nobody noticed — the edit doesn't undo the disclosure if the document was already transmitted or stored. The right move is to treat it as an incident.

Step 1: Stop distribution immediately. If the document is in a shared drive, remove it and log the removal time. If it was emailed, notify the recipients to delete and confirm deletion. Do not forward the document to anyone else for advice — that compound the disclosure.

Step 2: Notify your security officer or classification management officer in writing, with the source document identified, the destination document identified, the exact text copied, and the distribution that occurred. The incident report format varies by agency; the content doesn't.

Step 3: Cooperate with the damage assessment. The originating program office will evaluate whether the copied information requires notification to foreign partners, congressional oversight, or other agencies. Do not pre-empt that assessment by deciding the information "wasn't really sensitive."

Step 4: Expect the marking to be corrected in the new document and the source lineage to be restored. If the new document was properly classified all along but had a deficient authority block, that's a correction, not a disclosure.

The failure handling is not optional and it is not informal. Agencies track these incidents, and a pattern of unreported near-misses is itself a finding in the security inspection.

FAQ

Is copying classified text verbatim into a classified document plagiarism?

Not in the academic sense — the security framework doesn't care about attribution. It is a derivative classification act, and it's entirely proper if the new document inherits the source's markings and authority block. Verbatim copying into a classified container is routine and expected in intelligence products; the violation is the missing marking, not the copying.

Does it matter if I paraphrase instead of copying verbatim?

Paraphrasing changes the traceability, not the classification status. A paraphrase of classified information is still classified information and still requires the source's markings on the derivative document. If anything, paraphrasing makes the marking obligation harder to discharge because the text no longer visually matches the source, so the authority block has to carry the weight.

What's the difference between "derivative classification" and "declassification"?

Derivative classification creates a new classified document from an existing one. Declassification removes the protected status from information that no longer meets the standard. The two are handled by different offices and different authorities: derivative classifiers inherit markings, while declassification requires the original classification authority or its successor to make an affirmative determination.

Can I quote a classified document in an unclassified report if I cite it?

No. Citation neither authorizes nor sanitizes the disclosure. If the quoted text is classified, placing it in an unclassified document is an unauthorized disclosure regardless of the footnote. The correct path is to request a sanitized version from the originating office or to paraphrase in a way that removes the classified substance entirely, with the originating office's confirmation.

Who is legally responsible when a verbatim copy leaks?

The person who made the copy bears primary responsibility for the derivative document's markings, and the person who transmitted it beyond its authorized distribution bears responsibility for the disclosure. Original classification authorities retain responsibility for the source's marking correctness. In practice, investigators assign responsibility by act, not by role — the person who pasted the sentence is the derivative classifier.

What happens to a contractor who copies classified text into a proposal?

Contractors are bound by their facility security clearance and the National Industrial Security Program Operating Manual (NISPOM, 32 CFR Part 117). A verbatim copy into an unclassified proposal is an unauthorized disclosure under the NISPOM and typically triggers a report to the Defense Counterintelligence and Security Agency, a facility security review, and possible suspension of the cleared facility's eligibility to handle classified work.

How long does classified information stay classified, and does copying reset the clock?

Copying does not reset the declassification clock. The derivative document inherits the source's declassification date or event, which is set by the original classification authority under E.O. 13526 Section 1.5. A copy made in 2026 of a 2010 document with a 25-year duration still declassifies in 2035, not 2051, and the authority block must reflect the source's schedule.

This article is general educational information about classification and information-handling policy, not legal advice or an official interpretation of any agency's security regulations. Specific obligations vary by agency, contract, and jurisdiction; consult your facility security officer, classification management officer, or a qualified attorney for guidance on a particular document or incident.