The consequences of an attack and a classification of attackers
In order to be able to realistically estimate the strengths and weaknesses of attacks on the security of a smart card, it is important to first have at least a rough idea of the possible types of attackers. We can also use this information to help us devise defensive strategies and mechanisms. As a rule, typical attackers have one of two basic motivations. The first motivation is simple greed, while the second motivation is the desire for fame and status within a particular ‘scene’. These two motivations have different consequences for the system operator. An attacker who seeks a financial reward for his activities may take a certain risk by becoming a ‘card issuer’ in his own right,6 or he may attempt to blackmail the system operator. Both approaches can be combated using the usual judicial measures. If details of the attack become public, the reputation of the smart card system will be damaged. The worst damage to the reputation of a system operator occurs when a large number of cardholders lose money as a result of an attack. The reputation of a smart card system can be similarly damaged by an attack prompted by a compulsion to perform scientific research, rather than criminal tendencies. An attacker of this sort will consider his or her activities to be successful only if the results can be published in a suitable manner. The attacker is also under strong pressure to publish these discoveries as quickly as possible, since in this field, as is well known, being first is what counts. The end result is that the system operator, with little or no warning, is confronted with the publication of a detailed description of an attack on his system. Following this, the published attack is refined step by step by other interested parties and explained in terms that can be grasped by outsiders. The final blow comes when programs that carry out the attack in a fully automated manner are published on the Internet. In the spring of 1998, several GSM network operators found themselves confronted with a series of events similar to what has just been described. However, in this case the attack on the COMP128 cryptographic algorithm, which is used for A3/A8, did not have major negative effects on normal network operation.

There is a particularly significant aspect of this form of attack with regard to the attacker. He is regarded as the successful discoverer of a security leak, and thus as one of the ‘good guys’, and almost never need fear legal action as a result of his actions. The quintessential conclusion that can be drawn from these scenarios is that it ultimately does not particularly matter to a system operator whether an attack comes from a ‘good guy’ or a ‘bad guy’. In the case of a truly dangerous attack, the financial damage and the damage to the reputation of the system are most often rather large. In the worst case, the system must be shut down, all cards must be blocked and new cards that are immune to the attack must be issued. With a large system having several million cards in use, such a process can take more than half a year. The classification chart shows a classification of attackers based on the previously described aspects and practical experience. All types of attackers can be equally dangerous to a smart card system, but they have different capabilities and options. A typical hacker, for instance, has a moderate amount of system knowledge, good creative ideas and usually a similar group of friends. He normally does not have an extensive amount of equipment, and his financial means are also limited. However, if he is competent and employs a suitable approach, he can certainly obtain access to a large amount of processing capacity, for example by means of an Internet campaign. All insiders form a special class of attackers, under the assumption that they have very good knowledge of the system. They may have access to hardware and software components, and they may be aware of weaknesses in the system. As long as only single individuals are involved, they are equivalent to hackers in terms of their resources and options. However, since insiders are neither anonymous nor especially numerous, it is usually possible to identify the sources of their attacks.

The third class of persons who can be regarded as potential attackers is criminals. Although they usually do not have a high level of technical knowledge, they exhibit considerable energy when it comes to obtaining personal benefits (primarily financial) as a result of their activities. A potential source of attack that cannot be ignored in practice consists of academic institutions, such as universities and technical institutes, including their students and professors. They do not necessarily have special knowledge of particular smart card microcontrollers or applications, but they do have a large amount of generally useful knowledge. In addition, they have access to a large pool of qualified and inexpensive labor in the form of students and graduates, as well as adequate technical equipment in their laboratories. Many of these institutions also house a plentiful amount of processing capacity and highly motivated people with an experimental bent. A special class of attackers is formed by competitors. They normally have considerable technical knowledge, and some of them may have very sophisticated analytical equipment. Organized criminal organizations naturally represent a completely different level of attacks on smart card systems. They have sufficient financial resources to acquire all the knowledge and tools necessary for a successful attack, either commercially or by illicit means.