Are secure connections really that secure?

Computerworld - The little lock icon that appears on a Web browser window frame when a secure connection exists between a browser and a Web server may be lulling users into a false sense of security.

The reality is that secure connections, in which data is encrypted using Secure Sockets Layer (SSL) technology before being transmitted over the Web, is increasingly being used to hide and spread malicious code, according to a report from security vendor Kaspersky Labs.

The issue is certainly not new. Security analysts have for long warned about the possibility of hackers exploiting encrypted SSL connections to sneak viruses and other malicious code past firewalls, antivirus software and intrusion detection systems. But what's lending greater urgency to the issue now is the widespread use of SSL communications by banks, retailers, e-commerce sites and e-mail providers on the Internet, said Shane Coursen, a senior technical consultant at Kaspersky.

"A lot of people, when they go to a Web site and see the picture of the lock on their browsers, assume the connection they have with the server is secure" and pay little attention to the data being exchanged, he said. All that a secure connection is designed to do is to verify the identity with whom information is being exchanged and then use encryption to protect the information from being viewed or modified by a third party.

There is usually little validation of the content being transmitted during such sessions. As a result, rogue hackers can use the connections as a way to transmit and spread malicious code, including Trojan horse programs and e-mail worms on client systems and Web servers, Coursen said.

"There are misconceptions that technologies such as SSL indicate that a Web site is safe when, in fact, it is not," said John Weinschenk, CEO and president of security firm Cenzic Inc. "A Secure Sockets Layer function certifies that the server the browser is talking to is the genuine site and provides encryption of data being transmitted."

While the technology does have a valid use and does provide some level of security, it still allows hackers to exploit underlying applications, he said. "While large companies have taken significant measures to secure their sites, the fact remains that there are holes hackers can exploit, and personal information can be compromised unless proactive measures are taken."

Traditional antivirus tools and intrusion detection systems are inadequate because they are not designed to detect malware in an encrypted connection. So malicious data within secure channels can cause a significant amount of damage, Coursen said.

But options are becoming available to deal with the issue, said Pete Lindstrom, an analyst with Midvale, Utah-based Burton Group. Vendors of intrusion detection systems, for instance, offer tools that can intercept an encrypted data stream, scanning the contents for malware and then passing it along to the destination in encrypted fashion, Lindstrom said.

According to the Kaspersky report, however, the problem with that approach is that the data stream is modified -- meaning the Web server cannot verify the authenticity of a client and a client cannot verify the authenticity of a server.

Most antivirus vendors today also offer Web application plug-ins that allow for the content in secure connections to be inspected, Coursen said. But some applications such as Microsoft Outlook and Outlook Express don't work very well with the plug-ins, he said.

Read more about Security in Computerworld's Security Topic Center.