Most Antivirus Products Don't Stop New Attacks

LINDON, Utah -- A recent comparative laboratory test conducted by Independent Security Evaluators (ISE) found that most enterprise anti-virus (AV) products frequently fail to stop custom malware variants commonly used by hackers today. The ISE test, which reviewed the performance of major AV software products, determined that Avinti’s iSolation Server™ generally outperformed other solutions with a catch rate of 98.8% and that only two vendors, Avinti and Symantec, stopped greater than 90 percent of all malware variant instances used in the study. Other vendors total catch rates for all malware tested ranged from 53% to 11%.

ISE conducted a test that simulated a common technique of virus writers who wish to evade traditional AV signature-based products, morphing existing viruses to make them unrecognizable. The test was created without specific knowledge of any of the AV products tested. Thousands of virus variants were tested against the eight AV solution vendors, including Avinti, Symantec, McAfee, Tumbleweed, Ironport, Proofpoint, Trend Micro, and Barracuda.

“We found that overall [Avinti’s] iSolation Server was the most effective of the products we tested at discovering non-standard malware variants. Many of the other products in our tests were substantially less effective, allowing significant majorities of our malware corpus to pass undetected.” – taken from ISE Comparative Test Results study, May 18, 2007.

The lab test was commissioned by Avinti™, a developer of proactive e-mail security solutions, with the objective of determining if there is a quantifiable difference in other products and a benefit to using the company’s iSolation Server product to stop new malware and variants. The test used over 4,200 new malware variants that were passed through each of the products. The test not only proved iSolation Server’s unique approach stopped more malware than the other seven products tested, but also identified a major weakness in most AV products being sold today : a difficulty in stopping malware variants that are modified using very common tools.

“We feel like this test was very representative of the way new malware variants are developed,” said Avi Rubin, the President and co-Founder of Independent Security Evaluators. “The corpus of e-mail used for this test was created from common viruses and were modified using publicly available and home-grown tools similar to those used to get by traditional AV filters.”

Independent Security Evaluators