CERT defends vulnerability info restrictions
Full disclosure increases hack attacks
The long-debated question of whether software and network vulnerability data should be shared freely and immediately re-surfaced recently, as Carnegie Mellon University's CERT Coordination Center (CERT/CC), formerly the Computer Emergency Response Team (CERT), announced hooking up with a private-industry organization called the Internet Security Alliance to make its advance alerts and vulnerability database immediately available to members.
Several press reports have suggested that the publicly-funded CERT/CC will be making its database available to those willing to pony up anywhere between $2,500 and $50,000 annually for some manner of subscription service, but this isn't quite right. CERT/CC won't be collecting money directly in exchange for services; the costs cited are actually the ISA membership fees, which vary according to the size of the company seeking to join.
ISA member companies, which include NASDAQ, Mellon Financial Services, AIG, TRW and VeriSign, will have access to the CERT/CC database, or Vulnerability Catalog as it's called, via a secure distribution network, so long as they're willing to sign and abide by a non-disclosure agreement. Members will also receive advance vulnerability reports, and have the opportunity to share such information with one another in confidence.
Previously, CERT had maintained a policy of sharing software vulnerabilities immediately with the two US government bodies which support it, the US Defense Information Systems Agency (DISA) and the US General Services Administration (GSA), and with the software companies concerned. After forty-five days, during which the software vendor was assumed to be fixing its product, the group typically would make abstracts of the vulnerabilities public on its Web site.
CERT/CC says it will not use its public funding to offer new services to private companies, and notes that the cost of making the database available to the private sector will come out of the Alliance's membership fees.
That said, CERT still has its detractors among Internet security specialists, many of whom question the fairness of making current threat information which affects all Net users and systems administrators available to a select few, while everyone else must wait over a month for the free abstracts.
"The CERT venture will cost organizations upwards of $2500 per year for....services that are available for free or little cost elsewhere," Network Solutions (NSI) former Chief Security Officer Richard Forno writes in a nice rant posted at Infowarrior.org.
"For small companies without dedicated security staffs -- who don't know where to look for security vulnerability information elsewhere on the Internet and thus rely on CERT advisories as their sole security information -- not being able to participate in the ISA means that they are at a comparative disadvantage to larger companies that can afford such luxuries," Forno observes.
Indeed, a number of security-oriented sites do offer free vulnerability information, often as soon as it's reported. Though many make their full database available as a pay service, it's also true that essential information gets into the public domain notably faster than it does through CERT/CC, and can give a sysadmin a useful heads-up.
Those who advocate full and immediate disclosure maintain that by concealing a vulnerability long enough to enable the vendor to patch it, CERT/CC needlessly exposes the general Net population to exploitation through a hole of which they are blissfully ignorant.
On the other hand, publicizing a vulnerability before a fix is available does make it easy for would-be attackers to discover and exploit flaws they otherwise wouldn't have learned about.
"We do know that we usually see a large increase in attacks that exploit a particular vulnerability shortly after information about that vulnerability becomes public," CERT/CC Director Rich Pethia told The Register.
Pethia says that CERT/CC's own experience suggests that the danger from publicizing an un-patched flaw is greater than the danger from keeping it under wraps.
"While there are strong opinions on both sides of the debate, not all these opinions are supported by empirical data," he said.
"We believe, in the absence of data that demonstrates that attacks are in progress, that the lower risk approach is to publicly release vulnerability data once the technology vendors, or others in the technical community, have had at least an opportunity to find corrections or work-arounds."
So for CERT/CC the crucial question is whether it can be proved that full, immediate disclosure actually reduces exploitation in the real world, on the theory that forewarned is forearmed.
We don't pretend to know the answer to that one, but we'd be happy to hear from readers who think they do. ®