Windows Defender vs Top 100 Infostealers

[EricParker]

Yikes. The scary thing is that the real failure rate is probably even higher because samples on sample sites are more likely to have signatures.

[WiseFocus]

You might be the only one on YouTube who consistently tests Windows Defender and always keeps it thoroughly evaluated.

[cpuuk]

And yet MS want us to trust they can keep our data safe in Recall & Copilot (native App).

[Dosenwerfer]

Professional software engineer here. What really needs to happen is that major vendors adopt the UWP permission model so that Microsoft gets the support it needs to slowly force that model to users. Permission based app sandboxing is how mobile operating system have been doing it for over a decade and Linux is shifting towards this too with Flatpak. Every app should just be able to read and modify its own stuff, and when it wants to do anything else, it needs to ask explicit permission for that. The more permissions and kind of permissions it asks determines how risky it is and can be presented accordingly. The current UAC windows are a joke compared to a proper permission system like Flatpak.

It is crazy that in 2024 the standard way of doing things on Windows is still giving *every* executable unquestioned access to the entire system, browser files, etc. This is what makes info stealers possible in the first place. The solution is not trying to develop better malware detection. It is to fix the fundamental flaws of the system.

And yes, Microsoft did try this with Windows 8 and failed horribly, but that was mostly due to loads of other factors, most of which have improved dramatically since then.

[Cinibonswirl]

Windows defender needs a massive UI/UX overhaul. Knowing Microsoft it'll probably take 5 years

[sandmanmoderngamer]

I want to see malwarebytes vs. 100 viurs

[Musicbylzy]

Make the same test with Kaspersky and bitdefender as well please.

[Pirateking]

i wanna see if DefenderUI improves it a little bit since it did on ransomware test the last time you made a comparison with and without it

[goodjohnjr]

Not bad, considering that was Windows Security at default settings, I recommend that people use DefenderUI to set Windows Security to Aggressive settings. Leo needs to do this same test with that setup.

[HellDuke-]

On reddit these days, you will find the most common response to a user question about getting an antivirus to be "just use MS Defender, it's more than enough". I try to chime in that it's not that solid and might not be a good idea if you need a robust system, but the common consensus is that most users will never need more than Defender.

[BandwagonSmoke]

I think virtual operating systems are such a powerful security tool.

[scorit-zqyx]

You should do this again, and at the start show that you have the latest security updates and Windows updates installed. Also, should try this again with defender for Business.

[xDefcon]

It would have been nice to see a network live log of all the outgoing traffic made from the VM running the infostealers. It may be possible that Defender blocked the connections but did not delete the exe.

[IamLookingforWoody_________]

Honestly, your all videos are knowledgeable😊😊😊❤❤❤👏🏼👏🏼👏🏼.

[xavier]

People who don't work with statistics or data would have underestimated the 78% hit rate that Microsoft Defender has. This video cements Defender as the ultimate AV solution to date.

[PCFix]

Windows Defender doing 78% against 100 infostealers is honestly pretty solid for a free solution! 🛡 But, yeah, it's definitely not perfect—especially considering how many people on Reddit keep recommending it as “enough” for most users. I’ve noticed the consensus tends to be that most users won’t need anything more, but for a robust system, Defender still has a ways to go. 🤔 I’d love to see a follow-up test with DefenderUI set to aggressive, like mentioned in the comments. That could push the effectiveness further. 🔥 Also, comparing with Kaspersky or Bitdefender would be awesome, as those are top-tier! Keep up the great testing, PC Security Channel! 👏 #WindowsDefender

[LeoDavidson]

I'd also like to see "Windows Defender vs Top 100 perfectly innocent programs", because the false positives with Defender are ridiculous these days. Every time we renew our signing key (which has to be done every 3-4 years) it decides we're a virus for a few days.

[glenyoung]

Good to see someone doing theses tests without trying to sell a product so thanks for your post. I'm just wondering if it's possible to get a clear outcome when you are testing multiple products on a PC at the same time. Would the results be the same if you ran the full test with only one AV product installed at a time?

[kreature]

You are technically classing detections as deciding to delete a file.
If it kills the process it spawns preventing it from stealing anything I'd say it did it's job at a minimum level, even though file is still left on the system.
Without checking network traffic and processes spawned on execution of each it's very hard to say if it did anything to protect you at all.

[raven]

Appreciate your videos. Very informative as always.