Paid secur-o-ware is generally better than free, but not always by a lot
Antivirus tests that assess the effectiveness of security products from the moment users visit infected websites have exposed widely differing performances among the various anti-malware products.
The unsponsored tests by Dennis Technology Labs, which were run over a three-month period, revealed that the efficacy of paid-for anti-malware security suites varies widely, but that all of them beat Microsoft’s free product. The exercise also discovered that blocking malicious sites based on reputation or on the presence of at least one false positive (labeling a legitimate application as malicious) was an effective approach in all the tested products.
The researchers exposed test systems to infected websites that exploit vulnerabilities to drop malware on systems to look at how well web reputation and exploit detection built into modern security packages worked. “This means that we provide a complete environment, allowing the products to use a wide range of in-built technologies to defend themselves,” explained Simon Edwards, technical director at Dennis Technology Labs.
The researchers looked beyond file detection, behaviour analysis and on-demand scanners’ features and analysed how systems loaded with security packages behave when exposed to threats on the interwebs. The exercise – framed within the guidelines of the Anti-malware testing standards organisation (www.AMTSO.org) – looked at the efficacy of consumer, small-business and enterprise security suites.
Kaspersky Internet Security 2012 and Norton Internet Security 2012 both earned the highest “AAA” ratings for their consumer security software. BitDefender Internet Security 2013 and ESET Smart Security 5 both scored a “AA” rating”. Other consumer security packages put through their paces earned lesser rankings. Trend Micro Internet Security 2012 scored a “B” while AVG Internet Security 2012 only managed a “C”. McAfee Internet Security 2012 and Microsoft Security Essentials failed to achieve a passing grade.
Sophos Anti-Virus Business and Kaspersky Small Office Security earned the highest “AAA” rating for their small business products, with Trend Micro Worry-Free Business Security Services and Symantec.Cloud earning an “A”.
But despite a mediocre rating in other categories, Symantec Endpoint Protection was on its own with a triple AAA rating in the enterprise category. Only Kaspersky Endpoint Security for Windows, which earned an “A”, got anywhere close.
The Dennis Labs results went beyond on-access and on-demand scanning results and probed a wider range of blocking techniques found in modern anti-malware products. Using less usual (arguably more comprehensive) criteria also meant some well-regarded consumer security products that normally test well performed poorly.
“McAfee and AVG did relatively poorly in the consumer test because they were compromised quite a few times, and neutralised fewer threats than the better-performing products,” Edwards explained. “This could be because the web reputation systems were not as strong as those of the other products. That’s my best guess. The best ones simply blocked the sites and so did not have to grapple with recognising and terminating malware.”
The researchers’ results were also unusual in that they had different vendors winning each of the three of the categories. Edwards explained that tests can reveal difference between products, even from the same vendor.
“Sometimes this is because the settings have been changed for different markets,” Edwards told El Reg. “For example, a vendor may tune a business product to generate false positives less often than a consumer product.
“Vendors also include different sets of features and technologies in different products. Some may try out new engines or techniques in consumer products and then push these out to business products when they are sure that they are stable. Some take the opposite approach, trying out new technologies in their business products.
“There is another, more interesting reason, for the difference in effectiveness. There can be bugs in one product that don’t exist in another. For example, Kaspersky’s results differ because of one such issue,” he added. ®