VB2003: Virus experts debate bug names

"What's in a name?" That was the question computer virus experts were asking each other at a panel discussion of virus naming conventions at Virus Bulletin 2003 (VB2003), an annual gathering of the world's leading authorities on computer viruses, worms and malicious code that was held in Toronto last week.

Free IT resource Open Source Business Conference (OSBC) May 22-23, 2007 Sponsored by OSBC Free IT resource Virtualization Insights from Top Experts - Learn how virtualization gets real! Sponsored by Dell

Disagreements about what and how to name new worms and viruses have produced a confusing system in recent years in which antivirus companies often compete to be the first to "name" a new virus and in which the same malicious code often has two or more names assigned to it, experts agreed.

Moderated by David Perry, global director of education at Trend Micro Inc., the panel was made up of noted malicious code experts, who often name viruses, and antivirus representatives from large corporations, who struggle with the practical problem of defending against new computer virus threats.

Setting a jocular tone, and highlighting some of the comical by-products of the current virus naming system, Perry "named" each of the panel members in the style of current viruses. To the great amusement of audience members, panel member Shawn Campbell, global antivirus project manager for Ford Motor Co., became "Campbell/03" under Perry's guidance and Randy Abrams, release antivirus specialist at Microsoft Corp., was dubbed "W32.Abrams."

Behind the scenes, however, is a concern that the multiplicity of names assigned to high-profile viruses and worms may result in confusion that helps those viruses to spread.

Calling virus naming discrepancies a "major issue", Campbell said that IT experts within large organizations often end up boggled by differences between the name antivirus vendors are using to describe a threat and popular names for those threats that are propagated in the mass media.

Technical naming conventions are fine for virus experts, but they mean nothing to most employees and corporate executives who are more likely to remember names like "I love you" and "Melissa" than "VBS.LoveLetter.A" and "W97.Melissa.A." The result is that corporate antivirus experts waste valuable time and resources in an outbreak trying to reconcile the differences, Campbell said.

"Scientific names mean nothing to the public. The question is: 'What was it called when your manager saw it on CNN?' Maybe it's something you're already working on," he said.

At issue is a semi-official virus naming convention that dates back to the early 1990s and was developed by the Computer Antivirus Research Organization (CARO), a group of computer security experts.

Released in 1991 and occasionally updated since then, the CARO Virus Naming Convention set guidelines for what could and could not be used to name viruses and established a scheme that used attributes such as the type of threat (macro virus, Trojan horse), the platform affected and the family of threat, to come up with a name.

Speaking on behalf of the current system, panel member and CARO member Nick Fitzgerald said that the CARO system still works and that antivirus companies need to be careful that changes to existing naming conventions don't break proven antivirus engines that detect and thwart malicious code.

But others point to problems with the CARO Virus Naming Convention, especially the different results produced when legitimate antivirus researchers apply it to new threats.

To begin with, identifying and describing virus attributes is a subjective activity, according to Sarah Gordon, a senior research fellow at Symantec Corp. and VB2003 attendee.

In addition, modern "blended threats" often have many different attributes, resulting in monumentally long and complicated names that often resemble URLs (uniform resource locators) for Web pages.

Without any central repository of CARO names, confusion results, she said.

Microsoft's Abrams says that those discrepancies make his job harder.

As the person responsible for making sure Microsoft does not ship infected software to its customers, Abrams said that different antivirus products frequently identify the same malicious code differently, which complicates investigations of the threat.

A scientific naming scheme for viruses, akin to the current virus list maintained by The Wildlist Organization International or the Common Vulnerabilities and Exposures list maintained by Mitre Corp., would ensure consistency in the names that were assigned, Gordon and others agreed.

And with standard names, antivirus engines could be certified for compliance with the naming standard, Abrams said. Companies could have products decertified for not complying with the scheme, he said.

Despite the strong opinions, others expressed skepticism that the panel discussion and others like it will produce a consensus on what to call computer viruses.

Sounding a note of caution, panel member and virus expert Richard Ford of the Florida Institute of Technology noted that the antivirus community had been debating virus names for nine years, but still hadn't found an acceptable solution to the problem.