Three months of drafting, a paid review, then an outcome letter stating that your career episodes matched existing text. Many overseas-trained engineers first discover how the CDR plagiarism check actually works after rejection, not before submission. This guide walks through what Engineers Australia inspects, what a failed check costs in fees and visa time, and how to write career episodes that survive both text-matching and the newer AI-content layer.
The reassurance up front: original work written in your own voice passes comfortably. The trap is that “original” means more than a low similarity score, and most guidance online never explains the difference.
What Engineers Australia Counts as CDR Plagiarism
Engineers Australia’s Migration Skills Assessment (MSA) Booklet sets the originality requirement: every career episode and the summary statement must be your own work, written by you, in the first person. Two independent rules sit inside that requirement, and confusing them is the most expensive mistake an applicant makes.
The first is text-similarity plagiarism: copied or closely reworded passages from sample CDRs, shared templates, journals, websites, or another engineer’s report. The second is authorship plagiarism: someone other than you wrote the episode. These are enforced separately. A career episode that a paid service rewrites to a near-zero similarity score still fails the authorship rule, because you did not author it. No competitor page draws this line, and it is the one distinction that decides whether editing a ghostwritten draft saves you. It does not.
Copied Text from Samples, Templates, and Other Submissions
Lifting phrasing from a downloadable sample, reusing a template’s project narrative, or borrowing sentences from a colleague’s approved CDR all register as matched text. Reword the sentences, and you defeat nothing: the assessment tools and the assessor both read paraphrase, not only verbatim strings.
Why a Ghostwritten CDR Still Counts
Pay someone to write your career episodes, and you have breached the authorship requirement regardless of how clean the similarity report looks. The MSA Booklet frames the CDR as evidence of your own competency, so a third party’s writing misrepresents whose engineering judgment is on the page. Paraphrasing that draft afterward does not convert it into your work. The authorship bar, not the text-similarity score, is what Engineers Australia enforces here.
How Engineers Australia Runs a CDR Plagiarism Check
Turnitin is the tool named across the industry for Engineers Australia’s plagiarism detection, though EA does not publish its toolset, so treat this as widely reported rather than officially confirmed. Two layers matter, and they work differently.
Text-Matching and the Cross-Submission Database
The first layer compares your text against internet sources, academic databases, and, critically, a cross-submission pool of previously lodged CDRs. A March 2026 industry analysis of EA rejections describes assessors cross-referencing every career episode and summary statement against a global database, and singling out mosaic plagiarism, where fragments from several sources are stitched into one passage. That cross-submission pool is why a sample CDR that circulated online for years is one of the riskiest sources to borrow from.
Why Line-by-Line Paraphrasing Still Fails
Rewording each sentence keeps the underlying structure, the sequence of claims, and the technical framing intact, and that pattern is exactly what similarity engines are tuned to catch. Synonym-swapping a template produces text that reads as derivative to both the software and a trained assessor.
AI-Content Detection in 2025 and 2026
Turnitin’s AI writing detection is a separate module from its text-matching, and institutions configure the two independently. Curtin University, for example, disabled the AI-detection feature from 1 January 2026 while keeping text-matching active, which shows the layers are decoupled. For CDRs, the reported direction is the opposite. A 2026 migration-service report states that EA assessors are trained and equipped with detection software to identify the hallmarks of generative AI, and that an AI-generated submission can trigger a PIC 4020 fraud finding by the Department of Home Affairs on top of any EA sanction. A ChatGPT-drafted episode now carries two failure modes, not one.
What a Failed Plagiarism Check Costs in Time and Visa Progress
A failed CDR plagiarism check produces immediate consequences and typically a ban before you can reapply. Reported ban durations vary across the industry, commonly cited as 12, 24, or 36 months depending on severity. EA’s own site could not be verified for this figure during research, so treat the exact period as unconfirmed and plan for the longer end.
Then there is the money. Built from Engineers Australia’s public fee schedule effective 1 July 2026, here is the direct cost of getting back in the queue.
| Item (from 1 July 2026) | Fee (incl. GST) |
|---|---|
| Standard reassessment | AUD 1,034 |
| Admin reactivation | approximately AUD 132 |
| Combined minimum resubmission | over AUD 1,160 |
| Fast-track surcharge (optional, non-refundable) | AUD 396 |
| Appeal (refundable if successful) | AUD 704 |
[CDR Help Australia estimate: AUD 1,034 standard reassessment + approximately AUD 132 admin reactivation = over AUD 1,160 before any fast-track or appeal, based on EA’s published July 2026 fee schedule.]
The fees are the smaller concern. A ban of 12 to 36 months stalls the skills assessment that anchors your entire visa timeline, so the real cost is a delayed or lapsed visa pathway.
How the Finding Shows Up and What It Triggers
Where an assessor suspects non-original work, EA can issue an Additional Information Request asking for original project documentation, drawings, and calculation sheets, with a 30-day window to respond. Miss that window, or fail to produce evidence that you authored the work, and the file moves toward rejection. This is the step most applicants never see coming, because they assumed the process was a simple pass or fail on a similarity number.
Six Writing Patterns That Trigger a CDR Report Plagiarism Flag
1. Reusing a sample CDR’s project narrative. Even with names changed, the structure and phrasing already sit in the cross-submission pool.
2. Line-by-line paraphrasing of a template. Same skeleton, swapped words. Detection tools read the pattern, not just the vocabulary.
3. Mosaic plagiarism. Fragments paraphrased from several sources and stitched together. This is a named detection target, not a grey area.
4. Group-project double submission. Two engineers from the same project each describe the same site, the same design, and the same outcome. The cross-submission database can flag both, even when both wrote independently.
5. A writing service’s draft as your base text. This fails on authorship regardless of how much you subsequently edit.
6. AI-generated episodes. Uniform sentence length, an even and over-smooth register, and no first-person hesitation are the signals the AI layer reads.
How to Run a CDR Plagiarism Checker on Your Own Report
Check your report before you lodge, not after. A pre-submission CDR plagiarism checker catches matched text while you can still fix it. Grammarly’s plagiarism checker and Copyleaks both offer low-cost scans that flag overlap with web and published sources; Turnitin’s similarity report, if you can reach it through an institution, adds the academic-database coverage the consumer tools miss. None of these is EA’s own tool, so a clean consumer scan is reassurance, not a guarantee.
What Score Should Prompt a Rewrite
Here the honest answer matters more than a tidy number. A “5% threshold” circulates on many CDR sites as the pass mark, but Engineers Australia has never published an official similarity percentage, and that figure traces back to a service blog with no EA source. Chasing 5% misses the point: a ghostwritten episode at 2% still fails on authorship, and a genuinely original episode may read higher because it correctly cites a standard. Aim for authentic authorship and specificity, not a metric EA has never confirmed.
Writing Career Episodes That Pass Any Plagiarism or AI Detector
The writing that survives both layers of the CDR plagiarism check has one property in common: only you could have written it. Generic engineering description is what every sample, template, and language model produces, so it is what gets matched. Specific, first-person, site-anchored detail has no source to match against.
Compare these two sentences.
Generic and flaggable: “The drainage design was optimised to manage peak flow events.”
Site-specific and defensible: “I revised the culvert diameter at Chainage 2+400 from 600mm to 900mm after the hydrological model showed the original size would be overtopped in a 1-in-100-year event at the Toowoomba site.”
The second names a location, a parameter you changed, the analysis that drove it, and your decision. A text-matcher has nothing to match; an AI detector reads the hedged, first-person, idiosyncratic detail it does not generate. Write every competency claim that way, and both problems largely solve themselves.
Writing About Group Projects Without Matching a Colleague
When you and a co-worker both submit CDRs on the same project, describe your slice, not the project brochure. Anchor each episode to the calculations you personally ran, the trade-off you weighed when the model disagreed with the plan, and the specific judgment call that was yours to make. Two engineers who each document their own decisions produce two unmatched episodes; two who each summarise the same project scope produce a cross-submission flag.
For how these episodes are marked once they clear originality, see our breakdown of what EA assessors actually score and the CDR format each section must follow.
Questions Engineers Ask About CDR Plagiarism Checks
Does Engineers Australia Check Every CDR, or Only Suspicious Ones?
Every submission is checked. The plagiarism and AI screening run across all career episodes and summary statements as standard, not only on reports an assessor already doubts.
Can I Use a Published CDR Sample as a Structural Reference?
You can study a sample to understand the section structure and competency mapping, but do not lift its wording, project narrative, or phrasing. Samples that circulate online sit in the cross-submission pool, so borrowed text from them is among the fastest ways to match.
Will Paraphrasing a Writing Service’s Draft Pass the Check?
No. Paraphrasing lowers text similarity but does nothing about authorship, and EA’s policy requires that you write the report. A draft authored by a service remains a breach after you edit it, because the authorship rule is separate from the similarity score.
How Long Must I Wait to Reapply After a Plagiarism Rejection?
Industry sources report bans of 12, 24, or 36 months depending on severity, but EA’s exact reapplication period could not be verified from EA’s own documentation. Plan for a delay of at least a year, and confirm the current rule against the MSA Booklet before you reapply.
Does EA’s Check Detect AI-Generated CDR Content in 2026?
Yes, by industry accounts. EA is reported to apply AI-content detection alongside traditional text-matching, and to treat AI-authored episodes as an integrity breach, with a possible PIC 4020 referral to the Department of Home Affairs on top of EA’s own sanction.
Is There an Official EA Similarity Percentage Threshold?
No. Engineers Australia has never published a similarity percentage. The widely repeated 5% figure comes from a service blog, not EA, so treat authorship and specificity as the real bar rather than a number.
Engineers Australia flags more than plagiarism when it reviews a CDR. To see what else assessors reject reports for, and how to fix each issue before you lodge, read our full breakdown of the top reasons Engineers Australia rejects a CDR report, or work through the step-by-step CDR guide from the start.