More attack code surfaces for recent Microsoft security holes

Just days after Microsoft warned its customers about the release of code that can exploit a hole in its Secure Sockets Layer (SSL) library, new code that claims to exploit another recently disclosed hole surfaced on a French language Web site.

The computer code can be used by a remote attacker to trigger a buffer overrun vulnerability in the Local Security Authority Subsystem Service (LSASS), according to a message posted to www.k-otik.com. Microsoft released a patch for the LSASS vulnerability, MS04-011, April 13, along with fixes for the SSL problem and a number of other vulnerabilities.

The code was released on Saturday, according to the K-Otik Web site, which hosts the exploit. It was unclear Monday whether the exploit code worked, but notes attached by the exploit’s author say that some modifications to the code may be necessary before it could be used by a remote attacker to compromise Windows machines.

LSASS is used to authenticate users locally and also in client-server environments. LSASS also has features used by Active Directory utilities. An attacker who could exploit the LSASS vulnerability could remotely attack and take total control of Windows 2000 and Windows XP systems, according to Microsoft.

Unlike e-mail worms and viruses, no user interaction would be necessary to trigger the LSASS buffer overflow, according to Johannes Ullrich, chief technology officer at the SANS Institute’s Internet Storm Center.

The Internet Storm Center has not received any reports of the LSASS exploit code being used to compromise Windows systems on the Internet, he said.

Internet Security Systems Inc. (ISS) is also aware of the new code, but says it does not pose an immediate threat because it requires modification to work on computer networks.

“The exploit is unreliable and not for use in the wild,” said Neel Mehta, research engineer at ISS.

Not true for exploit code that targets the Microsoft SSL hole, which was released last week. ISS has seen a significant number of exploits using the SSL exploit since Wednesday, Mehta said.

Such activity is often a precursor to an exploit being used by a worm, he said.

The Internet Storm Center has received “a couple” of reports from organizations that had Windows systems attacked using that code, which leaves a unique signature in computer logs on compromised machines. The attacks were isolated, and do not appear to be linked to a worm or virus outbreak. However, there is evidence that malicious hackers have coupled the SSL exploit code with automated scanning tools, Ullrich said.

“It looks like, in some cases, all affected servers in part of a company got attacked. It seems like somebody picked a netblock (of network IP addresses) and started scanning those addresses and hitting all the affected systems,” he said.

On Thursday, Microsoft warned customers to “immediately install” MS04-011, citing “credible and serious” reports of the release of exploit code.

Any Windows XP, 2000 or Windows Server 2003 machine that runs applications that use SSL are vulnerable, including Microsoft Internet Information Services, Microsoft Exchange Server and third-party products, the company said.

ISS released an advisory on Friday that warned customers of the SSL exploit and cautioned that the severity of the Microsoft vulnerability was compounded by the fact that SSL is used to secure communications involving confidential or valuable financial information and that companies that use SSL must leave port 443, the port that is targeted by the exploit, open.

Systems that use SSL for secure communications are often “production critical” machines. Organizations take longer to patch such systems because of fears that applying the patch will interfere with critical services, Ullrich said.

Microsoft, ISS and other companies also have published work-arounds for the SSL vulnerability for organizations that cannot patch systems immediately, Mehta said.