What is the single most important way to improve endpoint security? According to Gartner’s Neil MacDonald, organizations should remove administrator rights from all users. Administrative rights on enterprise endpoints provide users with complete control over the device. These rights allow users to install software, change the Windows registry settings, change a wide variety of configuration files, and generally do whatever they want on the device.
Why are administrative rights a problem? Mainly because users might change the endpoint configuration or install unauthorized software. If unauthorized software installed by the user is benign, at most it would become a nuisance. But if unauthorized software is malicious, and installed under administrative rights, its impact can be devastating. In addition, since many Windows vulnerabilities that enable code execution do so in the context of the logged-in user, exploits might be able to execute without any restrictions on the endpoint. Therefore, we certainly agree that limiting administrator privileges for corporate end-users improves the organization’s security posture, but it’s not a panacea. Also, in today’s environments that support BYOC policies and ‘Consumerization of IT’, removing administrator rights is often unfeasible.
Removing administrator rights from the user does not prevent advanced malware infections. In his blog, Neil MacDonald says that removing these rights isn’t a “lockdown”; users will still be able to install software, drivers, ActiveX controls and more. This means that users will still be able to install potentially malicious files. Moreover, today’s advanced malware does not require user interaction or administrative rights to compromise an endpoint. Drive-by downloads, which exploit browser vulnerabilities and browser plug-in vulnerabilities, can infect the endpoint when the user simply views a compromised web-page (with or without administrative rights). This was the case in a recent Malvertising campaign recorded by Trusteer’s research team (see: Malvertising Campaigns Get a Boost from Unpatched Java Zero-Day Exploits). The attack utilizes a Java zero-day vulnerability (CVE-2013-0422) to automate the exploitation of the Java virtual machine. Embedded into ads that are displayed on legitimate websites, the exploit is able to automatically infect users with unpatched browsers when visiting these sites (without the users ever clicking on the ad).
Note that advanced malware can infect an endpoint when running under the context of either ‘administrative’ or ‘standard’ user rights, and in both cases, the malware can survive a reboot.
Drive-by downloads are a top attack method and they are growing in popularity with attackers¹. Today, drive-by downloads that are completely independent of user interaction pose a significant threat to enterprises because they are so hard to prevent. Attackers are taking advantage of the fact that many enterprises fall behind on patching endpoint vulnerabilities and are also exploiting zero-day vulnerabilities for which a patch is not available. We agree with MacDonald’s recommendation to use Application Control/Whitelisting to “lockdown” environments. Furthermore, we recommend that enterprises implement an Exploit Prevention Security Layer that uses an Application Control/Whitelisting technology to effectively protect vulnerable endpoint applications.