It was revealed recently that Sony's on-line services were the subject of another significant attack. This incident, however, did not exploit a vulnerability in Sony's security infrastructure so much as it highlighted the cascading effect of data breaches.

Rather than try to scale any fences or jimmy any windows, this attack used account holders' own keys to open the front door. According to a statement by Sony, the attackers tested a "massive set" of log-in credentials, consisting of pairs of user IDs and passwords, against accounts on three of its networks. Even though the "overwhelming majority" of the log-in attempts failed, they successfully breached about 93,000 user accounts. This indicates that the attackers used stolen log-in credentials, and did not resort to brute force or dictionary attacks.

How did the attackers obtain this trove of log-in information? Sony says it is "likely" they were stolen from elsewhere and not from its own networks, based on the low success rate. This may well be true, given the numerous incidents reported of late, some of which gave rise to our post referring to 2011 as The Year of the Breach.

If that scenario holds, it highlights the secondary effects of data breaches, and the relationship among user accounts on different on-line services. It has long been known that individuals often reuse the same username and/or password across multiple on-line services. As a result, if any one of those services suffers a breach that exposes its log-in information, corresponding accounts on the other services become open to the attackers. It is very much a "weakest link" situation.

This risk was also raised in the immediate aftermath of the data breaches at Sony this past Spring. The company initially reported the loss of unencrypted account passwords, which could have had the same cascading effect on its users' other accounts. Sony later stated that the passwords were in fact hashed. As we described at the time, "hashing" differs from "encryption," but storing passwords in a hashed form can be an effective way to keep an attacker from seeing or using the plain-text passwords of account holders. Password hashing is a known security technique that apparently was not in place at the "weak link" among the on-line services shared by those 93,000 users.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.