This week, the Mobile Antivirus Research Association, a collection of professors, authors and security professionals, announced it had "characterized' the first program to spread from PCs to a mobile device, a virus dubbed Crossover. In a rare occurrence in computer-virus circles, MARA appears to be the only organization to obtain a copy of the program--normally, such virus samples are sent by the creator to the major antivirus firms and shared among virus experts.

The exclusive access to the virus, and MARA's insistence that companies join its membership before being given access to the code, has antivirus companies up in arms.

"You have to go a long way back to find an analogous situation, where an antivirus group finds a virus and sits on the sample," said Mikko Hyppönen, chief research officer for antivirus firm F-Secure. "We didn't want to join their organization just to get a sample."

When F-Secure and other companies requested a sample from MARA, the organization sent back a legal agreement for membership. Among other rules, the document would have required that the company share its entire database of virus samples, Hyppönen said.

However, without the agreement, the Mobile Antivirus Research Association would not know if a new member would abide by the rules, said member and spokesperson Cyrus Peikari, the author of five books on security and the CEO of security firm Airscanner.

"Malware trading, which is illegal in many countries, should be done with a written chain of custody," Peikari said. "This is for the protection of all parties. Does it seem responsible for a legitimate company to send out viruses without some form of written agreement in place? I think some companies have let this slide for a number of years."

The debate over the virus sample has highlighted a rift between the more the conservative antivirus industry and a group of security researchers that do not adhere to the industry's stance against publishing virus code and associating with virus writers. Many security researchers believe that open disclosure of security vulnerabilities leads to better security. As those researchers begin to study viruses, worms and bot software, they argue that the same logic means the open discussion of threatening vectors for worms.

Already some security experts have moved toward the more open treatment of virus code. Last month, security researcher Kevin Finisterre admitted to creating the three versions of the OSX/InqTana worm and sending them to antivirus companies as a way to highlight weaknesses in Apple's operating system. In January, security consultant David Aitel revived the topic of beneficial computer worms, which he called "nematodes," and showed off research on how to conduct vulnerability scans of a network using the distributed features of such programs.

"In the general security community, it seems that the concept of 'security through obscurity' has long been discredited," Peikari said. "However, a small number of antivirus vendors are still strongly against re-publishing malware code in any form, even if it is already in the public domain--and readily accessible by anyone via free Internet download."