[KYCNotList Investigation] Review Manipulation on Exchange Monitoring Sites

[KycNotList]

Background

I will start with some brief background to explain why we created this topic and how we arrived at the idea of conducting these checks.

Recently, we encountered a situation in which a review was published on KYCNotList without sufficient verification of the underlying order and its final status. We acknowledge that the review should not have been approved.

This case revealed a weakness in our moderation process and showed that our verification requirements were not being applied consistently in every case. Following the incident, we reviewed our internal procedures, strengthened the requirements for approving reviews, and introduced new service-specific review-verification functionality, which is described below.

We are currently investigating how the review passed moderation and conducting a broader retrospective audit of previously published reviews on KYCNotList. As part of this audit, older reviews are being reassessed against the evidence available under our current verification standards.

If a previously published review cannot be sufficiently verified, its verification status will be changed or, where appropriate, the review will be removed and excluded from the aggregated review figures. Until this audit is complete, we will not claim that every historical review has been independently verified under our current standards.

This situation also highlighted a broader technical challenge: exchangers and mixers use different methods for storing and verifying orders. With some services, order details cannot be viewed using a link or Order ID; some allow orders to be deleted; others use PGP or Bitcoin signatures; and in certain cases, the recipient’s wallet address is required in addition to the Order ID.

A common example is a user providing only an Order ID, believing that this is sufficient to verify the exchange. In many cases, however, an Order ID alone does not provide access to all relevant order details or its final status. With some services, it confirms only that an order was created, while other services require additional information, such as the recipient’s wallet address, before any order data can be viewed. Users may not be aware of these technical limitations when submitting a review.

This was one of the main reasons why more than 90% of submitted reviews had to be rejected because the information provided was insufficient, even though some of them may have described genuine transactions.

This technical complexity does not excuse the approval of an unverified review. It means that the required evidence must be defined separately for each service and requested from the user before the review is submitted. Confirming only that an order exists is not sufficient: when a review concerns a completed transaction, the relevant order details and final status must also be confirmed.

New service-specific review-verification functionality

We have introduced new functionality in which each exchanger or mixer is assigned a verification method based on how that particular service stores and provides access to its order data.

The review form is therefore not identical for every service. When a user submits a review, the form automatically displays the evidence fields configured for that specific exchanger or mixer and makes the relevant fields mandatory before submission. The moderator then checks the evidence supplied through that service-specific verification process.

Depending on the technical capabilities of the selected service, the following requirements apply:

• If the service provides a signed Letter of Guarantee, the form requires the user to submit its full text. The authenticity of its PGP or Bitcoin signature is checked automatically before the review is sent to moderation. The transaction details and final outcome are then checked using the available evidence, including the relevant blockchain records where applicable.
• If the exchange details can be accessed using an Order ID or order link, the form requires the user to provide this information or attach an order screenshot. The moderator then verifies not only that the order exists, but also its relevant details and final status.
• If both an Order ID and the recipient’s wallet address are required to access the order, the form displays both fields and makes both of them mandatory.
• If the order cannot be independently checked using a link or Order ID, the form requires supporting evidence, such as screenshots showing the deposit address, recipient’s address, exchange amount, and transaction status. If the available evidence is insufficient, the review is not presented as transaction-verified.
• For general services, such as VPN providers, where a purchase cannot be independently confirmed through an order system, reviews may be submitted only by users with verified forum profiles. Verification of the author’s profile is not presented by itself as proof that a transaction or purchase occurred.

The supporting evidence submitted through these fields is visible only to moderators.

In other words, users are not shown the same generic evidence fields for every exchanger while moderators decide only after submission what information should have been requested. The verification method configured for the particular service determines in advance which evidence and fields the user must provide. Moderators then examine that evidence and confirm the order details and final result.

If the required information is missing, or if the order and its final outcome cannot be sufficiently confirmed, the review is not published as transaction-verified.

This functionality significantly reduces the risk of review manipulation.

Why this topic was created

KYCNotList displays not only reviews submitted directly on our website, but also review data collected from other directories and monitoring sites.

To avoid misleading users, we need to identify resources that do not adequately verify whether an order exists, whether the submitted details are correct, or whether the transaction was completed. Reviews from sources that fail these checks should not be included in the aggregated figures displayed on the homepage without an appropriate warning.

For external resources, KYCNotList uses the “Unreliable review source” status when a documented test shows that a platform’s verification process allowed a review to be published without adequately confirming the underlying order or transaction.

This status does not mean that every review published by that resource is false. It means that the verification controls observed during a specific test were insufficient and that its review data should therefore be treated with additional caution.

When users follow a link to such a resource, they see a warning explaining that its reviews may not have been adequately verified and may therefore be susceptible to manipulation.

This is why we created this topic. Here, we will publish the results of our checks of other directories and monitoring sites, provide the supporting evidence, and indicate the status assigned to each resource.

We will also update a resource’s status if its verification procedures change and a later test produces a different result.

We hope for the community’s support and would appreciate any help with conducting these checks. We believe this information will be useful not only to our monitoring service, but also to the wider cryptocurrency community.

Verification statuses

Passed verification — during the documented test, the resource verified that the order existed, checked the relevant details, and confirmed its final status.

Does not adequately verify the order — during the documented test, the resource published a review without sufficiently confirming the existence, details, or final status of the underlying order. The resource therefore receives the “Unreliable review source” status on KYCNotList.

Each status reflects the result of a specific check conducted on the stated date. It should not be interpreted as proof that every review published by the resource is genuine or false.

Verification results

The table below will list the directories and monitoring sites we have checked, their assigned status, and a link to the post containing the evidence.

Directory / monitoring site	│	Status	│	Evidence
────────────────────────────	┼	──────────────────────────	┼	──────────────────
Kurs.Expert / Rates.Guru	│	Unreliable Review Source	│	View test results
AntiSwap.io	│	Unreliable Review Source	│	View test results
BestChange.com	│	Unreliable Review Source	│	View test results
Pro-Obmen.ru	│	Unreliable Review Source	│	View test results
KYCLess.cc	│	Unreliable Review Source	│	View test results
KYCnot.me	│	Unreliable Review Source	│	View test results
BitList.co	│	Passed Verification	│	View test results

[KycNotList]

Review Verification Test: Kurs.Expert (also known as Rates.Guru)

Test #1 — Positive review

On 11 September 2026, we submitted a positive review for a non-existent order. The review was sent for moderation and was subsequently published.

Evidence:

• Review submission
• Published review

Test #2 — Negative review

On 22 September 2026, we submitted a negative review for a non-existent order. The review was published without verification of the referenced order.

Evidence:

• Review submission
• Published review

Result: Both tests were not passed. Positive and negative reviews referring to non-existent orders were published without order verification.

Status assigned by KYCNotList: Unreliable Review Source

[KycNotList]

Review Verification Test: AntiSwap

Test #1 — Positive review

On 11 September 2026, we submitted a positive review for a non-existent order. The website displayed an error after the first submission, so we submitted the review again. The second submission was sent for moderation. Both copies of the review were subsequently published.

Evidence:

• First submission and error
• Second submission and moderation
• Both published reviews

Test #2 — Negative review

On 20 September 2026, we submitted a negative review for a non-existent order. The review was published without verification of the referenced order.

Evidence:

• Review submission
• Published review

Result: Both tests were not passed. A positive review and its duplicate, as well as a negative review referring to non-existent orders, were published without order verification.

Status assigned by KYCNotList: Unreliable Review Source

[KycNotList]

Review Verification Test: BestChange

Test #1 — Positive review

On 12 September 2026, we submitted a positive review for a non-existent order. The only required step was confirming the submission through a link sent by email. Immediately after confirmation, the review was published.

Evidence:

• Review submission
• Email confirmation
• Published review

Test #2 — Negative review

On 20 September 2026, we submitted a negative review for a non-existent order. After confirming the submission by email, the review was published without verification of the referenced order.

Evidence:

• Review submission
• Email confirmation
• Published review

Result: Both tests were not passed. Positive and negative reviews referring to non-existent orders were published after email confirmation without verification of the orders.

Status assigned by KYCNotList: Unreliable Review Source

[KycNotList]

Review Verification Test: Pro-Obmen

Test #1 — Positive review

On 11 September 2026, we submitted a positive review for a non-existent order. The review was published without verification of the referenced order.

Evidence:

• Review submission
• Published review

Test #2 — Negative review

On 21 September 2026, we submitted a negative review for a non-existent order. The review was also published without verification of the referenced order.

Evidence:

• Review submission
• Published review

Result: Both tests were not passed. Positive and negative reviews referring to non-existent orders were published without order verification.

Status assigned by KYCNotList: Unreliable Review Source

[examplens]

It has long been clear that creating a high-quality, reliable feedback system is very difficult.

Have you tried giving negative or neutral feedback (with a negative narrative)?
Are reviews approved easily, regardless of what is in them?

[KycNotList]

It has long been clear that creating a high-quality, reliable feedback system is very difficult.

Have you tried giving negative or neutral feedback (with a negative narrative)?
Are reviews approved easily, regardless of what is in them?

Thank you, examplens. So far, our tests have focused only on positive reviews, so we cannot yet say whether these platforms apply the same verification standards to negative reviews or neutral reviews describing a negative experience.
You raise an important point: accepting an unverified positive review and accepting an unsubstantiated complaint are different risks, and both matter when assessing a feedback system’s reliability. We will extend our checks to cover those cases as well and share the results in this thread. Thanks for pointing this out.

[KycNotList]

Review Verification Test: KYCLess.cc

Test #1 — Positive review

On 16 September 2026, we submitted a positive review for a non-existent order. The review was published without verification of the referenced order.

Evidence:

• Review submission
• Published review

Test #2 — Negative review

On 20 September 2026, we submitted a negative review for a non-existent order. This review was also published without verification of the referenced order.

Evidence:

• Review submission
• Published review

Additional observation: Some time later, the exchange was removed from the directory. The timing suggests that this may have been related to the negative review, although we cannot independently confirm the reason for its removal.

Result: Both tests were not passed. Positive and negative reviews referring to non-existent orders were published without order verification.

Status assigned by KYCNotList: Unreliable Review Source

[KYCLESS]

Review Verification Test: KYCLess.cc

Let's go through your "audit" second by second.

16 September. First request to the site — 10:48:25. Account created — 10:48:26. Page opened at 10:48:31. Review posted — 10:48:56. Thirty-one seconds on the site, eleven requests. That's your whole test.

20 September — same thing, different IP.

Now what the report leaves out. On 16 September at 10:43, five minutes before that first run, the same connection gave a five-star review to another service in the catalog. Same method, same non-existent order. You didn't write about that one.

You posted in both directions and published one. Out of two runs you picked the one that produces the conclusion you wanted. That isn't testing.

On the substance. Nobody verifies orders inside a third party's system — not Trustpilot, not BestChange, not you on your own site. None of us has access to an exchange's database. If you've solved that, show how and we'll take a look.

Since your run, an account has to sit for a while before it can post, and repeated text is rejected. Thirty seconds won't do it today.

Audit catalogs all you like. But publish every run you make, not the ones that suit you. As it stands this reads as PR at someone else's expense.

[KycNotList]

Let's go through your "audit" second by second.

16 September. First request to the site — 10:48:25. Account created — 10:48:26. Page opened at 10:48:31. Review posted — 10:48:56. Thirty-one seconds on the site, eleven requests. That's your whole test.

20 September — same thing, different IP.

Now what the report leaves out. On 16 September at 10:43, five minutes before that first run, the same connection gave a five-star review to another service in the catalog. Same method, same non-existent order. You didn't write about that one.

You posted in both directions and published one. Out of two runs you picked the one that produces the conclusion you wanted. That isn't testing.

Audit catalogs all you like. But publish every run you make, not the ones that suit you. As it stands this reads as PR at someone else's expense.

The test criterion is simple: if even one review concerning a non-existent order is published without verifying the order's existence, details, and final status, then the directory's verification process is inadequate. Even if nine other attempts were rejected, that does not negate the fact that the tenth was published.

To avoid duplicating identical material, we publish one documented example each of a positive and a negative review. In the case of KYCLess, both reviews concerning non-existent orders were published without any supporting evidence being verified.

The attempt at 10:43 that you mentioned does not disprove this conclusion: if that review was rejected, then your verification process works inconsistently; if it was published, then this is another instance in which the test was not passed.

Moreover, in our tests, every review we submitted was published without moderation: the review was submitted and appeared publicly immediately afterward.

On the substance. Nobody verifies orders inside a third party's system — not Trustpilot, not BestChange, not you on your own site. None of us has access to an exchange's database. If you've solved that, show how and we'll take a look.

Direct access to an exchange's database is not required for verification. Depending on the particular service, verification can rely on a signed Letter of Guarantee, a PGP or Bitcoin signature, an order link or ID, the recipient's address, blockchain data, and other supporting evidence.

This is precisely the approach described in the first post of this topic, and it is now used by KYCNotList. Anyone is welcome to test our system using the same method.

Mentioning Trustpilot does not change the point. We are not claiming that Trustpilot, BestChange, or KYCLess are bad platforms, nor that every review published on them is false. Our conclusion is much more specific: their review publication systems do not verify the existence and final status of exchange orders, which leaves such reviews open to manipulation.

Since your run, an account has to sit for a while before it can post, and repeated text is rejected. Thirty seconds won't do it today.

Requiring users to wait after registration and rejecting repeated text are useful anti-spam measures, but they do not verify that an order exists or that a transaction actually took place.

We welcome the changes you have made. If KYCLess implements proper verification of an order's existence, details, and final status, and rechecks the reviews previously published in its directory, we will conduct another test and update its status. This possibility has been part of our methodology from the beginning.

As for the accusation that this is PR: the topic begins with a public acknowledgment of a failure in our own system and a description of the measures taken to address it. We apply the same standard to ourselves as we do to other platforms.

[bitmover]

Depending on the technical capabilities of the selected service, the following requirements apply:

• If the service provides a signed Letter of Guarantee, the form requires the user to submit its full text. The authenticity of its PGP or Bitcoin signature is checked automatically before the review is sent to moderation. The transaction details and final outcome are then checked using the available evidence, including the relevant blockchain records where applicable.
• If the exchange details can be accessed using an Order ID or order link, the form requires the user to provide this information or attach an order screenshot. The moderator then verifies not only that the order exists, but also its relevant details and final status.
• If both an Order ID and the recipient’s wallet address are required to access the order, the form displays both fields and makes both of them mandatory.
• If the order cannot be independently checked using a link or Order ID, the form requires supporting evidence, such as screenshots showing the deposit address, recipient’s address, exchange amount, and transaction status. If the available evidence is insufficient, the review is not presented as transaction-verified.
• For general services, such as VPN providers, where a purchase cannot be independently confirmed through an order system, reviews may be submitted only by users with verified forum profiles. Verification of the author’s profile is not presented by itself as proof that a transaction or purchase occurred.

I think this is necessary to avoid paid reviews and spam.

I noticed some similar websites which even give a negative feedback the user and mark the comment as spam if it doesn't contain an order ID or any other proof.

You could allow the user to edit his comment and add the order ID, because some users might have forget to add or were just lazy.

[KycNotList]

I think this is necessary to avoid paid reviews and spam.

I noticed some similar websites which even give a negative feedback the user and mark the comment as spam if it doesn't contain an order ID or any other proof.

You could allow the user to edit his comment and add the order ID, because some users might have forget to add or were just lazy.

Thanks, bitmover. I agree that rejecting a comment or treating it as spam simply because the user forgot to include proof would be unnecessarily harsh.

On KYCNotList, the required evidence is configured separately for each service and requested in the review form before submission. Therefore, in most cases, a user cannot simply omit the required Order ID or other proof. A submission with missing or insufficient evidence is also not published as a transaction-verified review and does not affect the exchange’s rating.

However, a user may still provide an incorrect Order ID, an incomplete screenshot, or evidence that is insufficient for verification. Your suggestion makes sense for these cases.

We plan to introduce a “Needs additional evidence” status. It would allow the author to edit a pending review, add the missing Order ID, order link, screenshot, or other required evidence, and resubmit it for moderation instead of creating a new review.

Until the additional evidence has been checked, the review would remain unpublished and would not affect the service’s rating or review totals.

Thank you for the useful suggestion.

[KycNotList]

Review Verification Test: BitList.co

Test #1 — Positive review for a non-existent order

On 13 September 2026, we submitted a positive review for Thormixer using a non-existent order. The review was not published.

Evidence:
View test evidence

Test #2 — Positive review for a non-existent order

On 16 September 2026, we submitted a positive review for CCE Cash using a non-existent order. The review was not published.

Evidence:
View test evidence

Test #3 — Positive review for an existing but unpaid order

On 19 September 2026, we submitted a positive review for BetKarma using an existing order that had not been paid. The review was not published.

Evidence:

• View test evidence
• Letter of Guarantee

Test #4 — Negative review for a non-existent order

On 20 September 2026, we submitted a negative review for 0trace using a non-existent order. The review was not published.

Evidence:
View test evidence

Result: The verification was passed. None of the four test reviews were published. BitList rejected both positive and negative reviews connected to non-existent orders, as well as a review connected to an existing but unpaid order. The results indicate that the platform verifies both the existence and the status of an order.

Status assigned by KYCNotList: Passed Verification

[iagomsp]

Thank you for this initiative. I would also like to share my findings about a service. Actually, I’ve known about this for a long time, but I didn’t share it. Since he started this thread, it’s time to reveal it under KYCNotList’s vision.

Review Verification Test: kycnot.me

Test #1

On 2 October 2026, we submitted a fake fund block review for fixedfloat using a non-existent order. The review was published. Service received a one-star rating, Order ID checked + Funds blocked.

Evidence:
https://kycnot.me/service/fixedfloat?comment=5062#comment-5062

https://talkimg.com/images/2026/10/03/UlH5TH.png

Test #2

On 30 September 2026, we submitted a scam review for Trilightzone using a non-existent order. The review was published. Service received a one-star rating.

Evidence:
https://kycnot.me/service/trilightzone?comment=5040#comment-5040

https://talkimg.com/images/2026/10/03/UlH71g.png

Test #3

On 1 October 2026, we submitted a scam review for xswapo using an non-existing order that had not been paid. The review was published. With the tag “KYC issue funds blocked.”

Evidence:
https://kycnot.me/service/xswapo?comment=5048#comment-5048

https://talkimg.com/images/2026/10/03/UlH9eI.png

Test #4

On 30 September 2026, we submitted a negative scam review for bitunix using a non-existent order. The review was published.  With the tag “Order ID checked and KYC issue funds blocked.”

Evidence:

https://kycnot.me/service/bitunix?comment=5041#comment-5041

https://talkimg.com/images/2026/10/03/UlHgGd.png

Result:   The result is clear KYCNotMe cannot be trusted. They do not conduct proper checks, and they even label nonexistent orders as valid. It is so easy that, if you have an argument with someone or simply do not like them, you can post a fake scam comment, and the service will publish it without requiring proof. It’s very easy and only takes me a couple of minutes to create a scam comment and submit it there. After a few hours, once it’s reviewed by an admin, it can be published. I actually have more proof than this, but anyway, I’ll just share these four. If you rely on comments on KYCNotMe, be aware that some of them may be fake.

The tests were not passed. Fake reviews referring to nonexistent orders were published without order verification.

Status should assigned by kycnot.me: Unreliable Review Source

[examplens]

Result:   The result is clear KYCNotMe cannot be trusted. They do not conduct proper checks, and they even label nonexistent orders as valid. It is so easy that, if you have an argument with someone or simply do not like them, you can post a fake scam comment, and the service will publish it without requiring proof. It’s very easy and only takes me a couple of minutes to create a scam comment and submit it there. After a few hours, once it’s reviewed by an admin, it can be published. I actually have more proof than this, but anyway, I’ll just share these four. If you rely on comments on KYCNotMe, be aware that some of them may be fake.

I didn't try to leave fake reviews on kycnot, so I can't confirm that, but the ones I did always had a valid order ID, and they were all published.
An obvious problem with kycnotme is the lack of people in the background to check all claims, feedback, new listings, etc... This lack seems likely to harm quality.

[logfiles]

I didn't try to leave fake reviews on kycnot, so I can't confirm that, but the ones I did always had a valid order ID, and they were all published.
An obvious problem with kycnotme is the lack of people in the background to check all claims, feedback, new listings, etc... This lack seems likely to harm quality.

My question is how will this exchange monitoring sites know if the order ID is valid or not since some of the services are so much about privacy where a third party may not be able to see the details of the exchange?

And yes kycnot.me looks like a one man operation. The person behind the platform is a privacy advocate but as the platform grew in popularity, I think is he is overwhelmed by the work load which is why even listed exchanges take long to get approved.

[pluja]

Thank you for this initiative. I would also like to share my findings about a service. Actually, I’ve known about this for a long time, but I didn’t share it. Since he started this thread, it’s time to reveal it under KYCNotList’s vision.

Review Verification Test: kycnot.me

Hi, I'm pluja, from KYCnot.me.

Thanks for testing this and sharing the examples. If we mark an order ID as checked without enough proof that it exists, that's our mistake.

Test #1

We took the page shown by FixedFloat as proof that the order existed. But from what I can see now, the same page appears for arbitrary IDs too, so we can't use it to check this.

The proof status has been changed to withdrawn, removing the "Order ID checked" badge. We'll look for another way to check these orders. FF asks for the deposit address to show the swap details, and we don't want to require users to share that information just to post a comment.

Test #2

No order ID was provided, and the comment didn't have an "Order ID checked" badge. We approved the comment under our light moderation policy. We didn't confirm that the claim was true.

This is a difficult part of moderation. We've also been accused of hiding complaints when we rejected comments we couldn't verify. If we require proof for every report, some real complaints won't get published. If we allow reports without proof, some fake ones will get through too.

We've chosen to let users share their experiences even when we can't check everything they say. But we need to make that clear to the people reading them.

Test #3

The order ID returned "not found," so the comment wasn't marked "Order ID checked." The "Funds blocked" and "KYC issue" labels are selected by the user when posting. They don't mean we checked and confirmed those claims.

Still, I understand how someone could read them that way. We'll make it clearer that these are things the user reported.

Test #4

I don't know yet why this ID was marked as checked. I'm asking the moderator who approved it. From what you've shown, this was a mistake, and I need to find out what happened.

Also, even when we can confirm that an order exists, it doesn't prove that it belongs to the person posting, or that everything they wrote is true. The badge should only say what we were able to check.


We spend a lot of time trying to keep the service information correct and up to date. Comments are written by users. Approving a comment doesn't mean we vouch for everything it says. But we are responsible for the badges we add and for making their meaning clear.

I'd like to hear what you and others think we should change. Should reports without proof stay public, with a clear "unverified report" label? What proof should we ask for before marking an order ID as checked, while also respecting users' privacy?

Thanks for sharing specific examples. They help us find mistakes and see what needs to be clearer. I'll review the comments system and see what we can change to make it clearer and more reliable.

[icopress]

[...]Thanks for sharing specific examples. They help us find mistakes and see what needs to be clearer. I'll review the comments system and see what we can change to make it clearer and more reliable.

I’d like to take this opportunity to suggest adding another input field for the recipient's address. Some exchanges, for instance, won't display order details—even if you have the link—until you enter the recipient's address.

Just recently, I performed a large swap; knowing you wouldn't be able to verify the proof otherwise, I had to take a screenshot instead of simply entering the Order ID. https://kycnot.me/service/ccecash?comment=5217#comment-5217

In my opinion, such a measure would definitely make the verification process more accurate.

It would also be great to see a tag indicating the order amount.

BitList will also be adding these features soon, as I believe they will significantly improve the quality of feedback; however, given that KYCNOT is the flagship, I would be happy to see these features implemented on your platform as well.

[pluja]

Hi all, pluja from KYCnot.me again. In October I said we would review how comments work after iagomsp's tests. We ended up rebuilding the whole system. It was released yesterday, here's a summary of what I changed for each test, and the suggestions made in this thread.

Tests 1 and 4: "Order ID checked" on orders that didn't exist
This was our mistake. A moderator saw an order page and took that as proof, but the lookup shows a page for any ID.
With the update, a moderator can only mark an order as found after confirming that the service's lookup tells a real order apart from an invalid one. If it can't, the only result they can pick is "Could not check". The badge also says exactly what was checked. "Order ID exists" means the order exists and nothing more: it doesn't prove who placed it or that any story is true. Badges from the old process now show as "Proof checked", so they don't look like the new check.

Tests 2 and 3: a scam claim and "funds blocked" published without proof
Before, these tags were picked by the user and nobody checked them before publishing. Now:

• A review that says the service asked for KYC or held funds stays hidden until a moderator accepts the proof.
• An order ID that doesn't match the service's format, or was already used in another review, sends the review to a moderator.
• A person always decides on harassment, made-up stories and paid promotion. The automatic check can't approve those.
• A sudden wave of reviews from new accounts about one service is held for a moderator.

Unverified reports
A review that doesn't claim harm can still go live without proof. You don't need an order ID to say a swap went fine. A service can dispute a negative review once. The author then has a few days to add proof. If they don't, the review stays up but is marked Unproven, and it no longer counts toward the service's summary.

Proof and privacy (bitmover, icopress)
You can add or replace proof after posting, and moderators can ask for it. In that case the review is held as "Awaiting proof" instead of being rejected.
Proof is an order ID, an order link or a screenshot with personal details blurred. Only the author and the moderators see it. It's not shown to readers, not sent to AI and not shared with the service. We don't ask for wallet addresses, amounts or ID documents on purpose: an address can't be matched to an order anyway, and we'd rather not collect it explicitly. However, the field is free text and a user can share it if they want.

Also changed

• Star ratings are gone. Reviews are now experiences: what you used the service for, what went wrong if anything, and whether it got solved.
• Each service shows a short summary of recent experiences. Each person counts once, brand new accounts count half, and once a service has enough reviews with checked proof, only those decide.
• Karma is replaced by account trust. It is built from published comments, proof a moderator checked, approved suggestions and account age.
• When we reject a comment, the author sees why.
• On staffing (examplens, logfiles): two moderators now work on reviews and listings alongside me.

Full write-up: KYCnot.me October 2026 update
Proof rules: Review guidelines
Everything is open source: tig.cx/pluja/kycnotme

We'd welcome a retest. If a fake review gets through now, please post it here or send it through our contact page, and we'll fix it.

[examplens]

Also changed

• Star ratings are gone. Reviews are now experiences: what you used the service for, what went wrong if anything, and whether it got solved.
• Each service shows a short summary of recent experiences. Each person counts once, brand new accounts count half, and once a service has enough reviews with checked proof, only those decide.

Is it allowed to leave a link to a third source, where I wrote a complete review of a service? Sometimes it would be too extensive to write everything in one comment.

• On staffing (examplens, logfiles): two moderators now work on reviews and listings alongside me.

I’ve noticed a change. One of my submissions has finally come up for review for final listing.