Article "Electronic contracts and software agents.
The conclusion of the electronic
contract through "software agents" A false legal problem? Brief considerations" by Giusella Finacchiaro; 1. Software agent is a computer program or electronic used autonomous characteristic and lack control by humans. It will not be directly control by subject, although the result of that use will be attributable to that subject. The problem of software agent is explanation not about regulative innovation even though issue legal of contract made by this software recently occur. The electronic data processing technology has been developed from “mobile agent” that know as “software agent” or “mobile code”. It move and adopts a different type of shape from the original type of program which decreases the traffic on the network when it goes to the data. This software is used in electronic commerce for management of large volume of data. The definition of software agent is unclear but anyway it is possible to say that it is a program of computer which is not managed by individual but can perform itself. In part of “agent” is similar to agency but civil law systems does not have an exact correspondence. The “software agent” means a “thing” in term of law. It does not have legal status and also cannot be an agent. The capacity of the “software agent” to begin legally binding contracts is utility to discuss about the “software agent” natural capacity to enter into lawful binding contract. In fact, the agent’s legal capacity is not required as well as the legal status is in question, not the capacity. However, the agent acting illegally in making contracts also becomes a legal problem. We know that the contract concluded via “software agent” is validity made. Although the contractual content has been predetermined and content has not contained in detail, it does not mean that it is unforeseeable. On the other, nobody entrust the manifestation of their own intent to an instrument of unforeseeable result The allocation of the risk of an error by the “software agent” in relation to a malfunctioning of this software needs to be addressed. This would be on the basic of the principles of auto-responsibility and reliance. The risk must be determined as having being accepted and borne by the subject who uses it. In addition, any defects that the “software agent” obtains will extend responsibility to those who have produced or developed it. The law effectively answers the requirement of regulating called “software agent” programs, framing it as a means of the will’s transmission within the limits of the law of contract. Thus, an ad hoc is not what is required. In conclusion, although the contract binding made by software agent will be appear a legal problem, the definition of this software required in this point more than regulate the new law. There needs to be correct definition combined with unambiguous with respect to the phenomenon requiring regulation. 2. The author claims that there has been an increasing interest in legal aspects of e-commerce contracts concluded via “software agents” and validity of such contracts. Despite some problematic cases taken place in the last century, contracts can be validly concluded by computer. The author clarifies what “software agents” mean. The technical designation of “software agents” is ambiguous. However, it can be defined as a new class of computer program, which operates for the benefit of the user in order to find and process the information, also to automate achievement of routine operations. “Software agents” cooperate with other program agents, domesticate a various type of architecture from the conventional programs in order to process data in the place where it must be found. The objective of these activities is removing from the individual superfluous traffic load. Nevertheless, the main aim of this article is specification of the legal problems “software agents” may produce. The emphasis is made on the legal status of the “software agent” and responsibility in the contracts concluded with the attendance of the “software agents”. The author suggests assuming possibility for the “software agent” to have legal status. Taking into account that the “software agent” is an artificial computer program, how could it be responsible for bad outcomes of the contract? Professor Finocchiaro leaves this question opened, although being sure that positive case does not facilitate the situation. For example, if we imagine the “software agent” has legal status, the fact that permission to the electronic action was given by the “software agent” independently must be shown. According to the Italian law, a contract concluded through the “software agent” is valid. The reason is – the "software agent" is understood as the will transference made by its owner in advance, which means that the program result is foreseeable. Consequently, talking of the responsibility for the damage which is caused by the software agent utilization, the producer of the software agent, not the user, bears the risk for the error. In the author’s opinion, this is the most efficient resolution because it removes the risk from the contract parties. Furthermore, this encourages “software agents” producers to pay more attention on the reliability of their programs. The author points out that the legal regulation for the “software agent” is sufficient, because it can be regulated by the contract rules as transference of the user’s will. There is no need in a new guideline for every new case. However, there exist International rules on e-commerce contracts which were domesticated into the national e-commerce legislations that have benefitted from them significantly. Obversely, UNCITRAL had confirmed the limitation of responsibility of the agents inappropriate. Commission suggests giving wider clarification of “automated computer system” to comprise the “software agents” into it and it must be adjusted that “software agents” can be used in contracts conclusion. The summary shows that the legal problems regarding “software agent” usage in contracts are the problems of its interpretation rather than regulation. If its meaning is certain, then the “software agent” can be safely utilized in the electronic contracts conclusion. (Finocchiaro, 2003) 3. It has been accepted that a contract can be validly made via computer technology even if the computer system operates autonomously. Although the computer system is not controlled by the subject directly, essential elements of the content in the contract can be considered as being predetermined, thus, intentional. Therefore, the results of the use of the computer system will be attributed to the subject even though the system is autonomous. One type of software composed that is considered in particular is a “software agent”. There are several definitions of the software agent mentioned. The first is technical-computing definition which states that; ‘it is a programme being used for electronic data processing and search the network to gather data upon which they must operate, reading their internal instructions for the priorities to be followed and respected’. Another is a legal definition which appeared in the Uniform Electronic Transaction Act provided by North American legislation. It is worded as ‘a computer programme or an electronic or other automated means used independently to initiate an action or respond to electronic records or performances in whole or in part, without review or action by an individual’. In other words, the software agent is a computer programme which has autonomous characteristics due to the lack of direct control by humans over the functioning of the programme. The software agent is a computer programme which is a “thing” in the eye of law. Therefore, the programme itself has no legal status and so cannot be an agent. Hence, there is no need to consider the capacity of the programme to enter into a binding contract. The software agent is a means of the contractual will’s transmission which has already been formed in the subjective dimension of its user. The contractual content can be seen as being predetermined; although it may not be set in detail but it is foreseeable. For example, the software agent is instructed to buy a particular book at the lowest price, with a top price of 15 Euros. Even if it is not certain that at what price the book will be purchased, there is a predefined limits of the price which is foreseeable. Hence, the will to conclude the contract exists and is predetermined. The responsibility for the defects of consent will therefore be attributed to the user rather than the software agent that serves the user. With regard to the allocation of the risk relating to a malfunctioning of the software agent, the risk is considered as being accepted and borne by the user. Furthermore, any defects contained in the software agent will extend responsibility to those who have produced or developed it. As mentioned above, the function of software agent seems to be able to be applied within the limits of the discipline of contract law; thus, there is no need for an ad hoc regulative innovation concerning the software agent. Therefore, it can be concluded that a contract can be validly made by means of software agents. 4. The issues regarding computer programs, also known as software agents, can be considered and solved with current rules and regulations. The only problem left is to clearly define what these software agents are, what behaviors do they have and are they entities which can be legally bound to specific contracts? The software agent is used for all of electronic ecommerce which controls a large amount of data. For example, when comparing the prices for a product on the Internet, a “software agent” is used to calculate the volume of information available. Although an electronic data processing technology has been developed on so-called “mobile-agents” which are recognized as “software agents” or “mobile code” the technical-computing definition of the software agent is still unclear. However, it is possible to express that the software agent is a computer program that performs and responds to electronic record by itself. Furthermore, it does not control by humans, designed with humans name and present the “agent” in term of anthropomorphic. As a result of this, software agent is indicated the “thing” in law because of program cannot take legal status by itself and cannot be an agent in fact. Discussing capacity of software agent to enter into legally binding contract is very useful even if no agent’s legal capacity is required. In case of software agent program, the legal status is in question, not the capacity. When contract is made by software agent, the case of false representation and concluded the contract unlawfully would be occurred. How could the “agent software” answer for the damage caused to the contracting third party? The proof of legal status is required to reply any defect in consent of attributable of the software agent. In addition, in order to give such legal status, the consent has not become an element of the contract predetermined by representations but was independently produced by the software agent itself. The contract conclude by the software agent is regarded as a contract validity made. In fact, the contractual has not been determined in detail does not mean that it is unexpected. On the other hand, who would trust the manifestation of their own will to an instrument of unforeseeable outcome? Nevertheless, the allocation of the risk of an error by the software agent must be determined as having being accepted and borne by the subject who uses the software agent as well as those who have produced or developed it. In conclusion, if ever there is legal problems relate to contract that made by software agent, the point is problem of explanation. Therefore, correctly definition is very necessary as well as associate with clarity the understanding with subject to the phenomenon requiring regulation. 5. Begins by saying that the only legal problem around contracting electronically is one of education and not regulation. The problems around creating a contract using a computer have been examined and it is now largely accepted that this is possible. The article looks in more detail at ‘software agents’ – used for electronic data processing and gathering information and used primarily in ecommerce to manage large volumes of data. There is not one consistent definition of a ‘software agent’ but the author of the article defines it as a computer programme with a lack of human control that has resulted in its autonomous features. There is a tendency to treat the ‘agent’ as something more human than it actually is. Researchers look to future situations when artificial intelligence has been created and these ‘agents’ do have human characteristics. A software agent it the equivalent of a basic computer user using a search engine or a programme that facilitates downloading. The programme itself does not have legal status and therefore it cannot be an agent in law. The article says that the capacity of the ‘software agent’ to enter a legally binding contract is not the issue, capacity is not in question legal status is. Why should a software agent be given legal status? The author argues that this will just complicate relations further by actually giving the ‘software agent’ the status of an ‘agent.’ Currently under Italian law a contract entered into by a software agent appears to be valid, viewing the software agent merely as a means of transmission. Who will bear the risk in a contract created by a ‘software agent?’ The author argues that the risk must assume to have been accepted by the user of the agent. The article concludes that ‘software agents’ are already sufficiently regulated and as a result no new legislation needs to be created to cover them. UNCITRAL believe that the model law is good enough at regulating the phenomenon without the need for overwhelming changes. 6. There are 8 sections in this article. The first section in this article is the contract’s conclusion through computers: the actual terms of reference. The next part is the software agent: a new legal problem? This section will explain clearly the definition of software agent. Then, the third section describes software agent in the future and how to manage with this problem. Subsequently, the legal status of the software agent, the contract concluded by the “software agent” and the allocation of the risk of an error by the “software agent” respectively. Next is the question “why no ad hoc regulation?”. Finally, this article will present the reader with a conclusion. The readers could understand briefly how software agent can make a contract safely. The article begins by discussing that the computer technology makes a contract and computer system might perform freely. Furthermore, computer system could predetermine the contract’s content. The next section describes the meaning of software agent which the North American Legislation in the uniform Electronic agent define as “a computer program or and electronic or other automated means used independently to initiate an section or respond to electronic records or performances in whole or in part, without review or action by an individual”. Finocchiaro also provides the synthesized definition as “software agent is a computer program in which the lack of direct control by humans over the functioning of the program has shaped its autonomous characteristics”. Additionally, software agent has many names such as intelligent agent, mobile-agents and mobile code. Moreover, it is used for evaluating the amount of the information so it is interesting for the person who studies artificial intelligence. According to the author in the third section, he said that the present tendency of software agent is anthropomorphism which is designed in a human name and used in human terms. The author claims that it should have some legal rules to solve the legal problem which happen from the software agent. Finocchiaro goes on the legal status of the software agent; it is just a computer program. Hence, it is only “thing” in definition in law and it has no legal status. There is no reason to give legal status to software agent but software agent can be premised in legal status as agent. However, it will effect in the contract, if software agent are interpreted as an agent in legal term. Then, he accepts that software agent can conclude the contract in the fifth section. The main contractual content has been prearranged while in the detail of the content has not been determined. Besides, this system is agreed by the user of software agent so the responsibility of the defects is to the user rather than to the software agent system. The section six insists that users who use software agent have to obtain risks. Obviously, the producer and the developer will also be liable with the user. Briefly, this part points out the allocation of the risk whereas the next part shows the regulation to control this computer system. He believes that there are enough regulatory frameworks for the software agent such as the work of the commission of the United Nations on electronic commerce; the model law on electronic commerce which is published in 1996 and the uniform rules on electronic signatures in 2001. UNCITRAL is another organization which regulates the software agent in both common and civil law systems. This organization legislate the model law on electronic commerce that article 13 paragraph 2 shows that sender is person who send message object from the computer system of the sender. The author also compares the definition between “software agent” and “agent”. He explains that “software agent” does not similar to “agent” because the theory of agency will limit agent‘s responsibility while software agent could not be used in this principle. Finally, Finocchiaro describes that the international organizations, especially in common law, emphasize that software agent could use safely in a contract.