That’s according the latest round of tests by AV Test, an independent IT security institute based in Megdeburg, Germany. The firm’s December 2015 results for Windows 8.1 devices compared 20 different anti-malware programs for protection, performance, and usability.
The results, outlined in the chart above, add findings based on three different ratings:
- Protection, which rates how effective a given application is at detecting viruses
- Performance, which rates how much a given application impacts your system while running
- Usability, which rates how many false positives a given application shows users after testing.
By these metrics, the best possible antivirus will detect nearly all actual malware, will not use up a lot of system resources, and will not see malware where none exists. The chart makes for an interesting overview, but download the complete report and you can dig more into the specifics.
For example: At this point, most applications catch over 99 percent of well-known malware. Many, including Kaspersky and BitDefender, caught absolutely everything. But Comodo was shown to be particularly weak here, catching only 96.1 percent of the known malware that should be easy to spot.
The real difference comes when you get to 0-day malware, which is yet to be documented widely but is possible to spot based on patterns. Nearly half the programs tested caught everything; most caught over 98 percent. ThreatTrack, however, caught a pitiful 78.3 percent. Also weak in this category: Microsoft’s Windows Defender, which is the protection included with Windows right out of the box, and which demonstrated only a 90-percent catch rate.
But there is such a thing as being too zealous: pointing out malware where none exists can confuse users and waste time. AhnLab’s software seems to enjoy falsely labeling legitimate software as malware, with 18 installed programs falsely identified. F-Secure also falsely declared nine legitimate apps as malware. Most programs didn’t do this, however, or only falsely labeled one app.
This all could change, of course — malware is a fast-moving world, and this month’s test results might not stand up next month. That’s why it’s good that AV Test is regularly running these tests, to keep the security companies accountable.
