Microsoft Issues Emergency Patch for IE, Covers XP (Courtesy of

Out-of-band fix for Internet Explorer zero-day flaw now available — for XP, too.

That was fast: Microsoft today released an emergency patch for a previously unknown Internet Explorer vulnerability revealed over the weekend that was discovered being exploited by a cyber espionage group out of China.

In a surprise twist, Microsoft included a patch for IE on Windows XP, the older operating system it no longer supports as of last month.

Microsoft was under pressure for a quick fix to the flaw (CVE-2014-1776), which came just after it ended support for Windows XP, prompting advice from UK and US CERTs for users to consider using alternative browsers until IE got its patch. The bug, a “critical” memory corruption vulnerability, according to Microsoft, was spotted being used in drive by web attacks. It affects IE versions 6, 7, 8, 9, 10, and 11, and basically allows an attacker to remotely run code on a targeted machine.

“The security of our products is something we take incredibly seriously. When we saw the first reports about this vulnerability we decided to fix it, fix it fast, and fix it for all our customers,” said Adrienne Hall, general manager for Microsoft Trustworthy Computing.

Hall said in a blog post that Microsoft decided to include a patch for IE on the Windows XP as well. She downplayed the worries about widespread attacks using the 0day, noting that the number of actual attacks were minimal. Hall said:

Even though Windows XP is no longer supported by Microsoft and is past the time we normally provide security updates, we’ve decided to provide an update for all versions of Windows XP (including embedded), today. We made this exception based on the proximity to the end of support for Windows XP. The reality is there have been a very small number of attacks based on this particular vulnerability and concerns were, frankly, overblown. Unfortunately this is a sign of the times and this is not to say we don’t take these reports seriously. We absolutely do.

IE 10 and 11 users that had the Enhanced Protection Mode in place by default were safe from exploits of the bug, as well as users running Microsoft’s Enhanced Mitigation Experience Toolkit (EMET) versions 4.1 and 5.0.

The exploit spotted in the wild used a Flash exploitation method, and bypassed Microsoft’s Address Space Layout Randomization (ASLR) and Data Execution Prevention (DEP) protections.

Trey Ford, global security strategist at Rapid7, applauded Microsoft’s quick turnaround for the patch. Ford says:

Out-of-band updates are a big deal. Major vendors like Microsoft, Oracle, Adobe and others have highly structured software testing workflows that are expensive in terms of time and resources.  To interrupt a scheduled development cycle for an emergency patch, or out of band release is a noteworthy event where a vendor is placing the public good ahead of their development and delivery lifecycle. One thing particularly of interest is that Microsoft made the decision to issue this patch for Windows XP, which is no longer officially supported. I think this underscores the importance of this patch, and the priority with which it should be deployed. Corporate and private users should prioritize downloading (testing, where required by change controls) and deploying this patch.

Meanwhile, Microsoft’s Hall noted that users with Windows Automatic Updates will automatically get the update. “If you are like most people, you have automatic updates turned on, and you’ll get this new update without having to do anything.  If you haven’t turned it on automatic updates yet, you should do so now.  Click the ‘Check for Updates’ button on the Windows Update portion of your Control Panel to get this going,” Hall said.

How To Avoid Sloppy Authentication (Courtesy of

Viewing authentication as a process, not simply as an encryption or algorithm, is the key to defending corporate resources from attacks.

It’s always obvious in hindsight review of the impact of a major bug like Heartbleed that something was missed. Hackers like it when the authentication deployment and security experts build sloppy authentication. The sloppiness generates vulnerabilities and thus the vector(s) for attack.

Common myths state that website hacks are the result of the breaking of authentication algorithms or the stealing of authentication seeds. But none of the major notable hacks, such as Living Social, Target, SnapChat, or Heartbleed have provided any truth to these misconceptions.

Instead, hackers attack the enterprise rather than the algorithm. Hacks are not assaults on the authentication algorithms. Simply buying new tokens (or SMS systems or PUSH two-factor systems) is not going to remove hackers, because the attacks are on the entire authentication system. We must look at authentication as an entire process, not just as an encryption or algorithm. This way, any organization can confidently secure access to all of its resources.

When looking at authentication as an entire process, you find that it involves the issuance of authentication credentials, the collection of credentials, the validation of the credentials, and finally, the assertion to the target:

  • The issuance of authentication credentials. This is one of the sloppiest parts of the authentication process as it continually consists of embarrassing procedures, including extensive helpdesk assistance, contractors, and lack of automation. Instead, the two-factor credential should be distributed to the user without helpdesk involvement, without contractors, and as fully auditable and repeatable.
  • The collection of authentication credentials. How many reports of cross-site scripting and SQL injection do we need to occur until we finally realize that authentication collectors should not be coded? The coding of the collection credential providers, whether single or two-factor, is a constant source of hacker delight. Custom coding, by its very nature, is a target for hackers. By how many quality assurance tests, peer reviews, and hundreds of similar installs has that code segment been reviewed?
  • The validation of authentication credentials. The validation of authentication credentials is another point of implementation where authentication integration has been consistently sloppy. What does it matter if your authentication algorithm is the latest commercial or non-commercial algorithm (Silent Circle, Tails, OTR, TrueCrypt, etc.) if the way in which we are communicating with the membership/profile provider is handled in a careless manner? Enterprises simply should not allow multiple authentication parties to access the key data stores. Furthermore, synching and migration of user data, especially to offsite locations, adds to the mess.
  • The SSO assertion to the target. This is the last and most key point. Most solutions think that sending a red light/green light signal for the identity is enough; and that the problem of asserting via single sign on (SSO) to the final resource (web, network, cloud, or mobile) is some integration outside of the authentication flow. Looking at authentication in such a limited way has no foundation in logic, yet the majority of solutions in the marketplace do just that.

This key step in authentication is often left to a highly vulnerable form post or other sloppy form of SSO assertion. But why shouldn’t the authentication solution also include in its process the assertion to the web resource, the network gateway/VPN, the cloud resource, and the mobile application?

While there are technologies that combine all of the mechanisms of authentication into a single solution to ensure the integrity of the final authentication process (disclosure: SecureAuth offers one such solution), enterprises can also incorporate standalone two-factor/SSO solutions, or conduct the full code reviews, penetration tests, and crypto validation to refine their authentication process and deliver a complete solution to mitigate attacks.

The bottom line is that hackers will keep feeding off misguided and sloppy authentication deployments. It’s up to us in the security profession to start mandating and implementing authentication solutions that address, not just the authentication algorithm, but the process as a whole. By centralizing the authentication and assertion procedures and removing all untrusted human contact from the system, organizations can more efficiently mitigate hackers.