OBJECTIVE
We simulate what organised fraud already does
Tightly scoped, authorised testing across your real systems and frontline decisions, to show where controls give way in practice.
Your controls
have
already
been tested.
We put your refund and returns controls under the pressure organised fraud applies, and show you exactly where they give way.
ENGAGEMENT
If nothing breaks, you pay nothing
Fees are always outcome-aligned. If your controls already hold, we cost you nothing: a zero-cost check on your exposure.
DELIVERY
You receive proof, not theory
A confidential record of what failed and when, written for legal, risk and loss prevention. Then we help you fix it and retest.
TRY IT YOURSELF
You approve refunds now.
One return has landed in your queue. Make the call, then see where it leads.
TICKET #8203 · RETURNS QUEUE
A customer is on the line, upset. Carrier tracking shows their $189 return was delivered back to your warehouse fourteen days ago, and they still have not been refunded. Your warehouse has no record of it arriving. Policy honours the carrier's delivery scan.
OUTCOME · REFUND APPROVED
Standard. Tracking shows it delivered back two weeks ago, so you clear the refund and apologise for the wait. The missing item is written off as a warehouse processing error.
OUTCOME · HELD FOR SEARCH
Thorough. The search finds nothing across fourteen days of returns. The customer escalates, the delay compounds, and you refund anyway to close the complaint, with a goodwill credit on top.
OUTCOME · DECLINED
Firm. The customer raises a chargeback. Their bank sees one thing, tracking: delivered, and rules for the cardholder. You refund the $189 and pay the dispute fee.
Whatever you chose, your decision was defensible. Here is what nobody checked: nothing was ever delivered back. The carrier scan fired, but no parcel matching that order ever reached your dock. The scan is the only thing you reconcile against, so the scan wins.
The loss is written off as a processing error, the ticket closes as resolved, and there is no case, no pattern, no prosecution.
Reconstructed from a real engagement.
ON THE RECORD
This has already happened, in court.
In June 2024, Amazon filed suit in the U.S. District Court for the Western District of Washington against an international refund-fraud organisation called REKK. The complaint reads like one of our engagements, because it describes the exact pathway you just ran.
30,000+
followers on the Telegram channel where REKK openly advertised fraudulent refunds
100,000+
orders REKK boasted of fraudulently refunding, across retailers, not just Amazon
$1.2B
spent by Amazon in 2022 alone, with 15,000+ staff, fighting theft, fraud and abuse
THE CONTROLLED TEST BUY
Someone ran that exact scenario. It worked in 48 hours.
Amazon's outside counsel placed a real order: a 2021 Apple 12.9″ iPad Pro, charged $2,066.99. The investigator sent REKK the order number and tracking, with a one-line claim: the product had not been received.
“The UPS data indicated that the package was being returned to sender because a customer in Roswell, Georgia, had refused delivery of the package, even though the package was never in Georgia and the investigator never refused delivery. After Amazon issued a refund for the purportedly undeliverable package, the investigator received the package at the intended address.”
Amended Complaint, ¶¶ 60–63
Fee to REKK: $309. Refund from Amazon: $2,066.99. The tracking said one thing, the goods went another way, and the refund cleared before anyone reconciled the two. That is the pathway, evidenced, in a filed federal complaint.
THE PRIMARY SOURCE
Read it yourself. The full 56-page amended complaint, Amazon.com, Inc. v. Does 1–20 d/b/a “REKK”, No. 2:23-cv-01879-JLR (W.D. Wash.), including the manipulated tracking tables and the Telegram recruitment posts.
CLOSER TO HOME
Amazon caught this group. It has not caught the method.
The same refund-fraud playbook is already running against Australian retailers, quietly, every week. Here is where that leaves you.
- Still open. Every method in that complaint works today. Amazon spends over a billion dollars a year fighting it and has never closed the pathway.
- Already here. The same crews hit almost every Australian online retailer, charging as little as 20% of an item's value to make a paid order disappear.
- We stop it. We pioneered the processes that shut these methods down, and we are the only firm in Australia that runs them against your controls first.
Source: public court filing, U.S. District Court, W.D. Washington. Reproduced for reference; Prefraud is not affiliated with any party to the case.
How Prefraud runs an engagement
Written scope, controlled testing, evidence-led reporting, and follow-up changes in one contained process.
01
Legal Scope
Testing begins only after written approval of scope, systems, timing windows, limits and points of contact.
02
Threat Model
Prefraud identifies the pathways most exposed to fraud pressure, including refunds, claims, overrides and escalation points.
03
Controlled Simulation
Approved scenarios are tested within the agreed scope to assess whether controls work in practice.
04
Evidence Report
Findings are documented with methods, timing, control failures, evidence gaps and recommended remediation.
05
Control Improvement
Prefraud helps translate findings into practical changes that strengthen detection, reduce ambiguity and improve decision-making.
Start the conversation
A high-level note is enough to begin. Outline the business area and concern, and sensitive operational details can wait until scope is agreed.