The recent attack on Sony has publicly paraded one of the predominant problems in incident response. While the immediate issue in incident response is of course the remediation of compromised systems and bringing these systems and IT services back online, it is entirely human to look for somewhere to cast blame. Let’s begin by fleshing out the “why” of attribution before engaging with the “how”.
Why Attribution?
For an affected entity, and for the security professionals working within it, attribution of the attacker(s) is only an ancillary concern. Certainly steps taken (or not taken) during the direct remediation of the breach are important and even essential to the later attribution effort, but priority of work must be on restoration of service. After service is restored, and the breach points corrected, the attribution process begins. Here are some reasons to perform attribution:
1. Attacker Centered Defense– If the attacker can be identified or at least have their tactics, techniques and procedures characterized, it may be possible to take steps to make their next intrusion more difficult and time consuming.
2. Legal/Governmental Retribution– Bringing criminal charges or taking other action against the attacker may deter future attackers, or may have deterrence against the current attackers if they are in a jurisdiction where criminal charges cannot be brought, but are vulnerable to intergovernmental or other pressures.
3. “Hacking Back”– This is an operation pregnant with a host of legal and ethical concerns. It may appear advantageous in the midst of an attack to attack back, but the technical reasons for doing so need to be looked at carefully. There are two scenarios where hacking back could reduce the impact of an ongoing hacking attack. One of these is a DDOS attack, where hacking back into the command and control systems of the attributed botnet may be a viable method to stem the attack. In the second scenario, stolen data belonging to the attacked entity has been definitively located (through attribution efforts) on a server, and then the entity performs a hacking operation to delete that data before it can be copied or moved. Both of these operations are patently illegal in the U.S. due to the Computer Fraud and Abuse Act. Both also carry a low probability of success or advantage for the attacked entity. If a DDOS operation is disrupted, other botnets can be easily employed by the attacker at short notice, or they may regain control of the compromised botnet. The command and control servers may also be innocent bystanders who are unknowingly playing host to malware. Taking down the servers of a hospital or local government or foreign military could have extremely serious human and geopolitical downsides. In the second scenario, the attacked entity would have to be extremely proactive to get the toothpaste back into the tube. In fact, it is so difficult it may as well be regarded as functionally impossible, and there is little chance of getting all the data before it is further disseminated.
Looking at the limited reasons for attribution and given especially the weakness of the third reason, it may be entirely reasonable for an attacked entity to decide not to pursue attribution. In the end, even having your attacker arrested will not undo the damage caused in the attack. The process of attribution is also not cost free. Additionally, any major actions taken on the basis of the attribution performed may only result in more reputation damage for the entity as the original breach and security failure is further publicized.
Why Attribute the Sony Hack?
Obviously an attack as devastating as Sony’s–which will probably involve hundreds of millions of dollars in damage to Sony Pictures Entertainment, in addition to a substantial erosion of their reputation and personal damage to many of their prominent employees–should be answered. To decide to not attribute the attackers would make their actions seem tacitly permitted, and probably contribute to further devastating attacks against other entities. This Public Order justification is the purview of the FBI, who has taken a lead role in the investigation of the attack. Sony, of course, probably wishes nothing more than for this incident to go away and may be leaning more towards active disinterest in attribution. It isn’t as if any future lawsuit will recover their lost capital, and their threats of legal action against those who published the more tabloid-friendly portions of their internal emails are at serious risk of the Streisand Effect.
Of course, SPE is not the only actor involved. There are cyber security companies and the U.S. Government at work here as well. Attribution is in the interests of cyber security companies for reputation and prestige reasons, as a proof of their aptitude and ability. On the government side attribution can be a marker of aptitude as well, however, retribution is also a significant influence. In the global and domestic political arenas strong attribution can aid in the pursuit of other interests. Attribution has been used by the U.S. Government previously to pressure the Chinese Government in regards to their cyber espionage campaigns, and charges have been filed in the U.S. against a number of their operators. So it comes to this. The primary reasons for attribution are political, as support for past and future actions and as an instrument of geopolitical pressure. The actions supported by attribution may include legislative efforts, international sanctions, and even electronic and physical attacks.
In Part Two of this work, having established the interests of the actors involved in attribution, we will look at the “How” of the SPE hack attribution as far as it is known publicly.