Symantec: Vista's Color-Coded Security Messages Can Be Spoofed

Windows Vista

The User Account Control feature in Windows Vista is designed to prevent individuals from making system changes that aren't authorized by their IT departments. The feature is supposed to prevent beguiled workers from installing software that could present a threat to their corporate networks.

If a user attempts such a change, he or she is greeted with an error message bordered in bright red informing them that the move isn't authorized. Notifications for supposedly innocuous changes not requiring administrator approval -- such as activating a driver or other component that is a built-in part of Windows -- are presented within a friendly, light-green border.

The trouble, according to Symantec security researcher Ollie Whitehouse, is that malicious code can "trick" Windows Vista into generating the green notification when it should be holding up the stop sign. "The user is presented with a UAC prompt that [falsely] claims that Microsoft Windows needs to elevate permissions ... not a third-party application," writes Whitehouse, on his blog on Symantec's Web site.

Whitehouse says the problem can occur when users try to activate a part of Windows Vista -- RunLegacyCPLElevated.exe -- that's supposed to make Vista compatible with older Windows Control Panel plug-ins. Files associated with RunLegacyCPL.exe can act as Trojan horses for malware that can then get written to unprotected areas of a user's hard drive after he or she gets the bogus green light.

id
unit-1659132512259
type
Sponsored post

"Microsoft is saying you should only see [the green dialog box] if the application is part of Windows," Whitehouse writes in his blog entry, which appeared earlier this week. "While it's true that RunLegacyCPLElevated.exe is part of Windows, it isn't true that the arbitrary DLL it loads and executes is," Whitehouse says.

Microsoft, in a best practices guide, concedes that Vista's color-coded warnings aren't a fail safe security measure. "The UAC prompts aren't a direct security boundary -- they don't offer direct protection," says Microsoft. "They do offer you a chance to verify an action before it happens."