MS to issue emergency patch for potent IE vuln
Researchers show exploits for IE 7 and 8
Customer Success Testimonial: Recovery is Everything
Microsoft will release an emergency update that patches the Internet Explorer vulnerability used to breach the security defenses of Google and other large companies.
The software maker has said that real-world attacks against the browser continue to be "very limited" and that they're effective only against version 6, which was first released in 2001. Still, researchers have determined that it's possible to exploit more recent versions using well-known techniques, causing the level of concern generated by the vulnerability to spiral since last week, when Google revealed that it 20 other companies were hit by highly sophisticated attacks that pilfered intellectual property and user data.
Independent researchers have since raised the number of victims to 34 and said source code was specifically appropriated.
"Given the significant level of attention this issue has generated, confusion about what customers can do to protect themselves and the escalating threat environment Microsoft will release a security update out-of-band for this vulnerability," George Stathakopoulos, the general manager of the company's Trustworthy Computing Security group, wrote Tuesday morning. "We take the decision to go out-of-band very seriously given the impact to customers, but we believe releasing an update out-of-band update is the right decision at this time."
Typically, Microsoft releases patches on the second Tuesday of each month so that IT workers who administer tens of thousands of PCs have an opportunity to test how the changes will affect their systems. The company releases security patches outside that schedule only in rare cases. Microsoft won't say when the out-of-band update will be released until Wednesday, at the earliest.
The previously unknown IE vulnerability was used to compromise PCs used by Google and at least some of the other companies that were hit by the attacks. While that exploit code was effective only on IE 6, researcher Dina Dai Zovi said he has developed proof-of-concept code that at least partially compromises machines running IE 7. Security firm Vupen Security went even further saying in a very brief advisory that it was possible to remotely execute code on machines running IE 8.
Microsoft said it is working to confirm those claims, but the findings are a good indication that with enough time, successful in-the-wild attacks on more recent platforms are likely unless an emergency patch is issued.
While the underlying null pointer reference flaw is present in all recent versions of IE, a safety feature known as DEP, or data execution prevention, makes it much harder to exploit the bug to remotely hijack machines when running more recent browsers and operating systems. The successful attacks against IE 7 and IE 8 incorporate well-known techniques to bypass DEP. So far, Dai Zovi has been able only to read sensitive files when exploiting IE 7 on Windows Vista, but he still can't modify system settings. He says he's close to being able to carry out more powerful attacks, including on IE 8, pointing out obvious limitations to the protection.
Still, he mostly agrees with Microsoft guidance that systems that use a variety of protections - including DEP, ASLR, or address space layout randomization, and protected mode - will be significantly protected from exploits targeting the vulnerability.
Computer users "should eat their vegetables and exercise regularly like people have been telling them for 20 years," Dai Zovi told The Register. "DEP is not an excuse not to eat your vegetables. These mitigations are speed bumps. They're not perfect, so it requires paying conscious attention." ®
COMMENTS
you're not listening...
[Moriarty] You're not listening Grytpyppe!
[Thynne] Hmm? Oh, er, I was just reading this advertisement on the back page of my suit. "Wanted: One Genuine Charlie for software development. Apply Microsoft."
So....
Will MS be withdrawin the "Upgrade to 7 or get hacked" advice?
KVM
Mostly because of the above example of active user cooperation in getting machines infected, I have virtualised the Windows machines in the office.
They now run the Windows boxen under KVM in snapshot mode, so users can only do so much damage by their compulsive habit of installing malware. As soon as they log out, the machines revert to a known-good state. The data is kept in a separate SMB server which amongst other measures, removes everything from it except for a pre-approved list of MIME types at 10-minute intervals.
And as a side effect of virtualisation, now the Windows sessions run on multiseat machines at about 25% the power and 35% the cost of the old setup. Not to mention that by snapshooting I have eliminated 99% of non-malware related maintenance :)

IT infrastructure monitoring strategies
What you need to know about cloud backup
Agentless Backup is Not a Myth
Top 10 SIEM implementer’s checklist
Customer Success Testimonial: Recovery is Everything