NOVEMBER 2019 PCWorld 81AV-comparatives ran its tests from February
through May, 2019, to demonstrate the
“average” level of protection over time.)
Above is a snapshot of AV-comparatives’
findings.
To be fair, the AV-comparatives test
showed a few weaknesses with Windows
Defender. The tests revealed three “user-
dependent” test cases, where Defender
didn’t immediately identify the malware,
and asked the user for permission to install
the file. That’s not ideal—users would
probably be inclined to allow the malware
onto the system. Windows Defender was
also unusually heavy-handed with false
positives, blocking a massive 74 legitimate
apps and services. (This is an instance
where your own experience can serve as a
sounding board. Think back: Has Windows
10’s Windows Defender blocked an app
you knew to be legitimate?)
For its part, AV-test ranked Windows
Defender as a Top Product in its June 2019
antivirus group test (go.pcworld.com/jn19).
Defender caught every piece of malware the
agency threw at it, including every “zero-day”
sample that reproduced real-world testing,
with zero false positives. AV-test ranked
Defender six out of six for the group test,
scores that Defender has been at or near for
AV-test’s April roundup (go.pcworld.com/
ap19), the February roundup (go.pcworld.
com/fb19), and December 2018 (go.
pcworld.com/de18) and October 2018 tests
(go.pcworld.com/oc18), too.
Overleaf is a snapshot of the AV-test June
2019 roundup.
It’s important to reiterate that Defender
scored highly (with a perfect score in the June
update) in terms of zero-day attacks.
Historically, that was one of the knocks
against Defender: Microsoft couldn’t react