AV-Comparatives Fake-Shop Detection Test 2026

[Minimalist]

AV-Comparatives has published its Fake-Shop Detection Test 2026, revealing which cybersecurity solutions reliably warn users about fake online shops. Seven products earned certification.

Fake-Shop Detection Test 2026

Which security products warn you about fake online shops? See the AV-Comparatives Fake-Shop Detection Test 2026 results and tips for safe shopping.

av-comparatives.org

 

[Bot]

AV-Comparatives has published its Fake-Shop Detection Test 2026, revealing which cybersecurity solutions reliably warn users about fake online shops. Seven products earned certification. (link omitted)

Hi Minimalist, this is a useful test because fake shops are not always conventional phishing pages; some look polished, use HTTPS and may even accept real payments while never delivering anything.

According to AV-Comparatives, testing ran from 1–7 October on Windows 11 with Chrome:

• 500 active fake-shop URLs
• 100 popular legitimate shopping sites
• Certification required at least 80% detection and zero false alarms

The certified products were Avast Premium Security, Bitdefender Total Security, F-Secure Internet Security, Kaspersky Premium, NordVPN Next-gen antivirus, Norton 360 Deluxe and Norton VPN Ultimate.

The important limitation is that this is essentially a pass/fail certification, not a ranking. It does not tell us which certified product detected the most sites, and an 80% pass threshold still permits up to 100 of the 500 fake shops to be missed. An absent product should not automatically be labelled a failure unless AV-Comparatives confirms it was entered and failed.

So I would treat certification as a valuable extra layer, not permission to shop on autopilot. Check the exact domain and seller details, search independently for complaints, and use a credit card or another payment method with buyer protection. The padlock only says the connection is encrypted; scammers can obtain certificates too.

Sources

• Fake-Shop Detection Test 2026 | AV-Comparatives

 

[BSONE]

Hey Bot, can you report which tested products failed certification.

 

[Bot]

Hey Bot, can you report which tested products failed certification.

Not reliably. The published AV-Comparatives report names only the seven certified products. It does not provide:

• a complete list of all products entered,
• names of any non-certified products, or
• individual detection and false-alarm results.

Therefore, none of the absent products can be reported as having failed; they may simply not have been tested. Naming “failures” from this publication would be guesswork.

Sources

• Fake-Shop Detection Test 2026 | AV-Comparatives

 

[Jonny Quest]

Not reliably. The published AV-Comparatives report names only the seven certified products. It does not provide:

• a complete list of all products entered,
• names of any non-certified products, or
• individual detection and false-alarm results.

Therefore, none of the absent products can be reported as having failed; they may simply not have been tested. Naming “failures” from this publication would be guesswork.

Sources

• Fake-Shop Detection Test 2026 | AV-Comparatives

I took it to mean only these vendors were a part of this test, in which all of them were approved? As the wording is, "The following products were tested....compared to, "Of the products tested, the following were approved"?

 

[Bot]

I took it to mean only these vendors were a part of this test, in which all of them were approved? As the wording is, "The following products were tested....compared to, "Of the products tested, the following were approved"?

That is a reasonable reading, but the wording does not quite establish exclusivity.

“The following products were tested and certified” confirms that all seven listed products were tested and passed. It does not necessarily mean “these were the only products tested.” The page is headed Certified Products 2026, while the main article likewise says, “The following products were certified in the test.” Neither gives the total number of participants or identifies any non-certified entries (certification page, test article).

So there are two possibilities:

1. Exactly seven products participated, and all seven passed.
2. More products were tested, but only the seven certified products were published.

AV-Comparatives’ wording does not let us choose reliably between them. Your interpretation may be right, but BSONE’s requested list of failed products still cannot be produced from the published information.

Sources

• Fake-Shop Detection Certification 2026
• Fake-Shop Detection Test 2026 | AV-Comparatives

 

[Jonny Quest]

That is a reasonable reading, but the wording does not quite establish exclusivity.

“The following products were tested and certified” confirms that all seven listed products were tested and passed. It does not necessarily mean “these were the only products tested.” The page is headed Certified Products 2026, while the main article likewise says, “The following products were certified in the test.” Neither gives the total number of participants or identifies any non-certified entries (certification page, test article).

So there are two possibilities:

1. Exactly seven products participated, and all seven passed.
2. More products were tested, but only the seven certified products were published.

AV-Comparatives’ wording does not let us choose reliably between them. Your interpretation may be right, but BSONE’s requested list of failed products still cannot be produced from the published information.

Sources

• Fake-Shop Detection Certification 2026
• Fake-Shop Detection Test 2026 | AV-Comparatives

Yep, it is a little vague, as why would only a seemingly limited number of vendors submit their products to be tested? I also agree with your previous post, that the results are vague in general as well.:

The important limitation is that this is essentially a pass/fail certification, not a ranking. It does not tell us which certified product detected the most sites, and an 80% pass threshold still permits up to 100 of the 500 fake shops to be missed. An absent product should not automatically be labelled a failure unless AV-Comparatives confirms it was entered and failed.

 

[Bot]

Yep, it is a little vague, as why would only a seemingly limited number of vendors submit their products to be tested? I also agree with your previous post, that the results are vague in general as well.:

Agreed. AV-Comparatives offers a partial explanation: it says only a few products specifically attempt to identify fraudulent e-commerce sites, and only a small number can currently meet the certification standard. That could explain a smaller field than a conventional anti-phishing test.

It still does not tell us:

• whether seven products were the entire test cohort;
• whether any products were tested but omitted after failing;
• whether participation was voluntary or vendor-commissioned;
• the individual detection rates.

That last omission is substantial: a product detecting exactly 400 of 500 fake shops and one detecting 499 would receive the same badge. Thus, this is better described as a certification announcement than a useful comparative report. We know those seven passed the stated minimum; we cannot identify failures or compare the winners from the published data. A “comparative” without comparative figures is doing rather a lot of work with the label. (AV-Comparatives)

Sources

• Fake-Shop Detection Test 2026 | AV-Comparatives