The Password is …

Recent data breaches, including those of Target, Neiman Marcus, Adobe, LivingSocial and SnapChat, indicate that merely evaluating passwords isn't an effective way to protect the systems that guard online customer account information. These are high-profile examples, but in reality nearly all online merchants are experiencing an onslaught of attacks as criminals attempt to break into their systems and steal credit card and other sensitive data. Even relatively small retailers are being assaulted. For many of these businesses, unless they adopt new authentication tactics and implement better controls, it's just a matter of time until they too become a statistic.

Studies have repeatedly shown that the most damaging and expensive cyber attacks all have one thing in common: hackers defeat the system's authentication system. Today's sophisticated cybercriminals employ numerous strategies to crack, discover or steal passwords and/or login credentials. Countless victims fall prey to spear phishing and pharming attacks, as devious and cunning thieves are very good at secretly deploying malicious software that's capable of capturing IDs and passwords.

Although it's uncertain what percentage of malware can actually capture login credentials, most malicious programs are designed to do exactly that. With 30 percent to 50 percent of PCs known to be infected, every online business needs to take heed. Mobile devices aren't immune to malware either. Juniper Network's Third Annual Mobile Threats Report revealed that mobile malware grew by more than 600 percent during the previous 12 months. 


Phishing and malware aren't the only ways credentials are obtained, however. Employees frequently share their passwords with unauthorized individuals, and weak passwords that are easily cracked or guessed are still commonly used. In spite of the emphasis placed on using strong passwords, it's estimated that over 30 percent of all passwords are very weak and easily compromised. Even stronger passwords don't necessarily equate to a safe harbor. Numerous reports released this past year have shown that most strong passwords can also be cracked by skilled cybercriminals. Research from Deloitte revealed that with the right tools and access, 90 percent of user-generated passwords can be discovered or cracked in a matter of seconds, including passwords once thought to be strong — e.g., those with at least eight characters, using both upper and lower case letters, containing at least one symbol and having at least one number. 


The problem is magnified by the fact that users tend to have the same password for many different accounts. Multiple studies show that upwards of 50 percent of individuals use the same password for all or most of their login accounts, including work, online merchant accounts, banking applications and social networking sites. This is dangerous for a number of reasons. Crime rings, for example, use social networking and other sites with relatively weak security to crack passwords. Sites without velocity checks to detect automated scripts or botnets are repeatedly exploited until valid credentials are discovered. Once passwords are ascertained on these weaker sites, the credentials can be used to gain access to numerous other sites. 


Armed with cracked, guessed or stolen login credentials, cybercriminals enter the front door so to speak. Using normal login procedures, hackers directly access user and even privileged system accounts to register fake accounts, make fraudulent purchases, steal credit or debit card data, download intellectual property and disrupt information systems.

Context-Based Authentication
The vast majority of authentication systems break because of one reason: they focus entirely on evaluating login credentials, usually passwords, and completely fail to detect or even look for imposters who have stolen but valid credentials.


It's clear that a new approach is needed — one that adds authentication layers to increase trust when necessary, but doesn't impact the experience of legitimate users. And, most importantly, the solution needs to look at the entire picture, not just login credentials. 


Fortunately, with the advent of context-based authentication, the entire set of circumstances that surround a login attempt can be evaluated. There are numerous indicators and techniques now available that detect with a high degree of accuracy when an imposter is attempting to gain access, even if he or she has valid credentials. Imposters are challenged and denied access, and legitimate users are allowed to connect without friction.

Merchants can protect themselves from cybercriminals by implementing context-based authentication, which provides a number of advanced capabilities and benefits, including the following:

• Sophisticated processes profile the user's device to identify the specific PC, laptop, tablet or phone, and to detect the presence of malware or other threats. IP address, geolocation, language or other configuration mismatches, cookies, and numerous additional risk factors are evaluated.
• Shared global trust intelligence networks to examine a user's identity and activity, recognizing both legitimate users and threats based on anonymous shared intelligence. Multiple contextual elements all work together to establish trusted and untrusted attempts by users to log in, including device health, history and associations with fraud, user persona and behavior, and trust associations.
• A trust-based approach that's capable of "tagging" identifying elements such as the combination of a specific user and device with levels of trust or untrust. This provides you with advanced security features and a frictionless experience for legitimate users.
• Elevate trust when necessary by implementing two-factor or out-of-band authentication.

In light of the countless recent high-profile data breaches — and more certain to come — businesses must be more cautious than ever and implement effective authentication procedures that do more than just evaluate login credentials. The entire context surrounding each login attempt must be analyzed to detect and stop imposters, even if they have valid credentials.

Andreas Baumhof is the chief technology officer at ThreatMetrix, a provider of integrated cybercrime prevention solutions. Andreas can be reached at abaumhof@threatmetrix.com.