(DM n. 10 del 23 gennaio 2001)
Anno 2001 - prot. 2001122198_003

Parte: I
1.1 Programma di Ricerca di tipo: interuniversitario

Area Scientifico Disciplinare: Scienze giuridiche (65%)
Area Scientifico Disciplinare: Ingegneria Industriale e dell'informazione (20%)
Area Scientifico Disciplinare: Scienze storiche, filosofiche, pedagogiche e psicologiche (15%)

1.2 Durata del Programma di Ricerca: 24 mesi

1.3 Title of the Research program

Software agents and electronic commerce: comparative and economic analysis of legal rules

1.4 Coordinatore Scientifico del Programma di Ricerca

(cognome) (nome)  
UniversitÓ degli Studi di BOLOGNA FacoltÓ di GIURISPRUDENZA
(universitÓ) (facoltÓ)
(settore scient.discipl.) (Dipartimento/Istituto)

1.5 Responsabile Scientifico dell'UnitÓ di Ricerca

(cognome) (nome)  

Professore ordinario 30/09/1959 PSCGNN59P30C352J
(qualifica) (data di nascita) (codice di identificazione personale)

UniversitÓ degli Studi di TRENTO FacoltÓ di GIURISPRUDENZA
(universitÓ) (facoltÓ)
(settore scient.discipl.) (Dipartimento/Istituto)

0461/881855 0461/881899
(prefisso e telefono) (numero fax) (E-mail)

2.2 Settori scientifico-disciplinari interessati dal Programma di Ricerca

2.3 Keywords

Testo inglese

2.4 Base di partenza scientifica nazionale o internazionale
The expression 'intelligent agents' or other equivalent terms (i.e. software or electronic agents) are used to call, in generic sense, a typology of software capable of independent action in complex contexts.
This type of software has been integrated in the electronic commerce.
From the economic analysis of law point of view, the impact of the intelligent agents on the electronic commerce can be analyzed both from the consumers and from the enterprises perspectives. As for the former, the possibility to reduce the cost associated with the products search would have to increase the transparency of the market and to guarantee a sharp price cutting. It can also be predicted that many functions of human intermediation are destined to being replaced from the automated systems. The possible reactions from the enterprises must be taken in account and carefully considered. Some vendors could decide to block the access of the agents; the same consumers may decide to stop using the agents, in order to avoid to delegate the purchase choices completely. One of the main problems consists in understanding if the current or future technology will be able to avoid illicit commercial praxes, equivalents to those currently realized with the use of standard contractual terms.
As for the enterprises side, the intelligent agents can be used for the inner management and the relationships between the several divisions of the firm, or towards the outside, for the personalization of the marketing policy towards the virtual costumers. The main problem is the creation of systems of interaction between agents who assure a efficient outcome of the automatic negotiations. Therefore, the next challenge, in this field, is represented from the development of ability to adaptation to situations of uncertainty and to complex environments in continuous change.
It can be observed that the current configuration of this kind of software and its possible future development add a new dowel to the crisis of the conceptual system on which the classic regulation of the market is based (assumed, in a broad sense, as the set of rules that concern contracts and competition).
To the typical factors of the digital age, synthetically indicated as dematerialization, delocalization and technologization, that already render impossible the simple transplant in the cyberspace of the cornerstone ideas on which is based the regulation of the real world, the intelligent agents seem to add a new one: the depersonalization.
The problem seems to rest with the worsening of a phenomenon which seems to be co existential to the development of Internet, that is the splitting between the real and the web identity, which has been already at the center of well known controversies (paradigmatic in this respect seems to be the case of the domain names conflicts). However, it seems certain that the intelligent agents represent a change of substantial quality with respect of this already known scenario. This change seems due to two reasons: the autonomy enjoyed by the agent with regard to the subject (physical or legal person provided with an electronic identity), and the invisibility of the same agent.
The subject who uses the agent does not have a direct and constant control of its functions, and the subject who receive the results of the agent activity does not perceive - or it cannot perceive - the effects of the same action. This consideration holds true for a set of different relationship. When the agent's user is the enterprise and the consumer receives its actions, or when it is the other way around. But also when the two subjects of the relationship are homologous (enterprise to enterprise). The depersonalization scenario stands out, then, more clearly, when the relation is established between two or more agents. Moreover, it can be presumed that for the third subject called to decide on the relation (judge, arbitrator, other type of authority) will not always be easy to lead back the agent and its activity to one determined user (physical or legal person). Currently, the use of the intelligent agents regards, therefore, three fundamental moments of the market functioning: the search and the supply of information (being it from the side of the consumer, or from the side of the enterprises), the negotiation and the conclusion of the exchange of products (informative and not).
From the legal point of view, what has been said highlights the following basic issues:
1) if the use of the agents has legal relevance;
2) if a contract entered through the use of agents is to be considered legal and, if so, which limits can be assigned to its validity;
3) if and how can be attributed to the agents' users the liability arising from the use or from the bad functioning of intelligent agents;
4) which instruments can be employed in order to avoid the setting of antitrust practices through the use of intelligent agents.
The traditional legal categories and rules in the field of agency, formation of contract, precontractual and contractual liability, consumer protection and antitrust, do not seems apt to solve the problem just listed. The observer may add to this partial conclusion that if one looks on a global scale to the rules that are currently being elaborated to handle this new set of problems the European and U.S. approach seem to be very different in this respect.
While the 2000/31 CE directive on the electronic commerce is completely silent on the new legal issues posed by the intelligent agents, the recent American model law -- approved in July 1999 by the National Conference of Commissioners on the Uniform State Laws, under the title of Uniform Computer Information Transaction Act (UCITA) - envisages a set of rules specifically designed to regulate the use of these agents. While it is true that the two kind of norms have different nature and a different applicative field, it must be stressed nonetheless that once again the U.S. legal system has been extremely rapid in setting rules for the use of the I.A., at least with regards to the contracts that deal with the transfer of information.
The American approach has been ready to give the broadest legal relevance to the use of I.A.
Nonetheless, such a sharp choice does not put all the doubts aside, while it raises serious problems of coordination with the traditional common law rules. It must be considered anyway that the model law and its first state enactments are way too recent for setting up today an analysis on the ability of this regulation to endorse the UCITA's goal: fostering the safest electronic business environment possible, and reducing to the maximum the legal uncertainty related to its development.
On the European side, the law-making silence is an even more meaningful data. It does not seem that it should be interpreted as a willful renounce to regulate the IA issues, but as a momentary difficulty, expression of the intrinsic complexity of the European legal process. As a matter of fact this latter is traditionally subjected to a strong competition with the competitor (the American one) and must reconcile the different 'souls' (common law - civil law) of the internal legal systems.
Following from the above considerations it seems clear that the intersection between I.A. and electronic business, as in other field of the law of the digital age, points out in the direction of new institutional structures, legal categories, and operational rules. The conceptual standpoint when the law faces new issues is always based on the traditional legal models and on the first generation rules, that the jurist must filter using the analytical tools of the comparative approach. The EAL offers then another theoretical instrument that can be used in order to evaluate which, among the starting models, can furnish the most suitable base for developing the new rules that will make possible a valuable and fair use of intelligent agents in the electronic business.

2.4.a Riferimenti bibliografici

AA.VV., Intelligent software agents: foundations and applications, Berlin, 1998;
AA.VV., Le applicazioni dell'intelligenza artificiale negli intermediari finanziari, Maspero D., Rossignoli C.( eds.), 2000;
AA.VV., Software agents for future communication systems, Hayzelden A. L.G., Bigham J. (eds.), Berlin, 1999;
COOTER R., ULEN T., Law and economics, 3. ed., Reading, Mass., 2000;
DAVIDSON S. J., ANDRESEN K.A., UCITA and other U.S. Laws in an International Perspective, 600 PLI/Pat 553 (2000);
FRIEDMAN H.M., Securities Regulation in Cyberspace, New York, 1998;
FROOMKIN A.M., Article 2B as Legal Software for Electronic Contracting - Operating System or Trojan Horse?, 13 Berkley Tech. L. J. 1023 (1998);
KERR I. R., Spirits in the Material World: Intelligent Agents as Intermediaries in Electronic Commerce, 22 Dalhouisie L.J. 190 (1999);
LEROUGE J. F., The Use of Electronic Agents Questioned under Contractual Law: Suggested Solutions on a European and American Level, 18 Marshall J. Computer & Info. L. 403 (1999);
LESNICK L., MOORE R. E., Agenti di ricerca, Milano, 1997;
LOCKERBY M. J., UCITA: The Uniform Computer Information Transaction Act, 7 Richmond Journal of Law and Technology 2 (2000);
NIMMER R. T., Through the Looking Glass: what Courts and UCITA Say about the Scope of Contract Law in the Information Age, 38 Duq. L. Rev. 255 (2000);
NIMMER R. T., UCITA: A Commercial Code for the New Commerce, 617 PLI/Pat 635 (2000);
PASCUZZI G., Internet (diritto privato), Digesto civ., Torino, 1999, vol. appendice , 531;
POGGI C. T., Electronic commerce legislation: an analysis of European and American approaches to contract formation, 41 Virginia Journal of International Law 224 (2000);
SMED S., Intelligent software agents and agency law, 14 Santa Clara Computer and High Technology Law Journal 111 (1998);
VULKAN N., Economic Implications of Agent Technology and E-Commerce, 109(453) Economic-Journal 67 (1999);
ZILIANI C., Gli agenti intelligenti: trasparenza dei prezzi e potere degli acquirenti su Internet, Industria e distribuzione, 2000, 71.

2.5 Descrizione del programma e dei compiti dell'UnitÓ di Ricerca
The present research project is aimed to analyze, with the interdisciplinary tools of the comparative approach and economic analysis, which are the more suitable legal models for developing rules apt to govern the interlacement between intelligent agents and electronic commerce.
From the legal standpoint this phenomenon raises the following pivotal problems:
1) if the use of the agents has legal relevance;
2) if a contract entered through the use of agents is to be considered legal and, if so, which limits can be envisaged to its validity;
3) if and how can be attributed to the agents' users the liability arising from the use, or from the bad functioning, of intelligent agents;
4) which instruments can be employed in order to avoid the setting of antitrust practices through the use of intelligent agents.
The research will be articulated in two phases.
The first one (of the duration of twelve months) will be dedicated to the collection and the analysis of the doctrinal literature, the relevant norms and the case law on the subject. Initially, focus will be placed on the gathering of the economic literature aimed at describing the impact of the intelligent agents on the e-business market dynamics. Subsequently, it will be taken in consideration the legal literature - mostly American -- that has begun to inquire the interlaces between intelligent agents and electronic commerce.
Later on, it will be collected the norms -- like the recent North American law-model, approved in July 1999 from the National Conference of Commissioners on the Uniform Been Laws with the title of Uniform Computer Information Transaction Act (UCITA) - that take specifically in consideration the use of the intelligent agents in the electronic commerce field. Eventually, the existing case-law on the subject will be gathered and indexed.
A careful examination of the gathered legal material will enable to carry out a theoretical reconstruction apt to extrapolate the main legal models of reference, tracing their origin from a given legal system (common law or civil law, Europe or U.S.). At the end of the first phase it will be produced a report on the scientific results of the work.
In the second phase (of the duration of twelve months), it will be proceeded to evaluate, through the legal comparison and the economic analysis, the various legal models emerging from the first phase of the research project. The final goal of the project is to produce an hyper textual e-book that will contain a monographic analysis of the research topics and a digital data-base where the relevant bibliography, norms and case-law will be made easily available to the searcher.