Internet Dispute Resolution Mechanisms and Applicable Law
Catherine KESSEDJIAN <firstname.lastname@example.org>
(1) What is the Internet for the Private International Lawyer?
(2) What consequences, if any, must be drawn from the fact that messages exchanged on the Internet may travel at random through numerous countries?
(3) When Primus, a company incorporated under the laws of Germany, contracts with Secundus, a company incorporated under the laws of New York, via the Internet, what law can they choose? We will assume the contract to be a simple sale of goods.
(4) Can Primus and Secundus decide in the contract which court will hear their potential dispute arising out of the contract? Can they choose arbitration? Can they provide for any alternative dispute resolution mechanism? Can they decide that the arbitral tribunal will meet virtually via the Internet?
(5) Secundus sold Primus' goods to end-users, via its Web site (the end-users access Secundus' Web site from all over the world). Some of Primus' goods are defective. Can end-users sue Secundus and, if so, where? Can end-users sue Primus, and, if so, where?
(6) Secundus is dissatisfied with Primus' products and is losing a lot of goodwill. Primus is a company known worldwide. In order to oblige Primus to act, Secundus uploads on its Web site criticisms on Primus' acts and on its management style. Primus considers these criticisms to be defamatory. What forum has jurisdiction to hear the case Primus vs. Secundus? Would the answer be different if Secundus were incorporated under the laws of Switzerland?
(7) In order to better know its client base, Secundus created a data bank, including many personal data on individuals who access its Web site. The French CNIL (Commission Nationale de l'Informatique et des Libertés) happens to learn that Secundus' data bank includes information on 15,000 French nationals and that Secundus will use these data in a manner violating French law. Can the CNIL enforce French law against Secundus? If so how?
(8) Finally, if the answers to questions 1 to 7 are not satisfactory, can we think about "regulating" Internet through an International Treaty? If so, what kind of a Treaty should be prepared? Which Organization(s) should take the lead? Is this urgent?
The Internet is yet another means of communication through which all kinds of legal relationships can be formed and maintained. The legal relationships formed through the Internet are not basically different from those formed through other means of communication. They are mainly contracts and torts. They can be assimilated to contracts between absentees and distant torts.
What makes the Internet specific, however, is the fact that it is transnational in nature and by essence. It is the global power of an Internet communication which makes it unique. A company opens a Web site to advertise its products or services. The Web site can be accessed from every country in the world where there are telecom and computer connections. Up until now, technology does not allow that company to forbid access to its Web site by "visitors" domiciled in a certain jurisdiction.
Because of this inherent transnational nature of the Internet, the impact of each legal relationship formed through Internet will inevitably affect several countries at the same time. We cannot assume (and it will be a long time until this actually happens) that the laws of those countries affected by the transaction are similar or even approximate (even within the European Union, the approximation or uniformization of private law is far from being complete). Thus, there must be rules allowing the different systems to coexist and to decide which law has the right to apply to a specific transaction. These rules are called "conflict of laws rules."
Private International Law deals also with the jurisdiction of courts and the extent to which private parties have the right to elect a court or choose to have their dispute resolved by arbitration or any other alternative method (mediation, conciliation, mini-trial, for example).
By answering the questions raised by the following hypotheticals, we will show private international law rules in the working.
2. What consequences, if any, must be drawn from the fact that messages exchanged on the Internet may travel at random through numerous countries?
Probably no consequence must be drawn from the fact that the messages exchanged via Internet travel through one or more countries. This is because, first, the route followed by the message is completely at random and, second, the means of communication used to complete a transaction is not relevant, in private international law, to determine which national law applies or which courts have jurisdiction.
On the contrary, much more important are the location of each of the parties, where the transaction must be performed, possibly the language of the contract, and the currency used for payment under the contract.
3. When Primus, a company incorporated under the laws of Germany, contracts with Secundus, a company incorporated under the laws of New York, via the Internet, what law can they choose? (We assume the contract is a simple sale of goods, Primus being the seller and Secundus the buyer.)
As for any other contract, a contract formed via the Internet can be submitted by the parties to a set of rules or a national law they may consider appropriate to their contract. This is known as the principle of "party autonomy." The rules chosen need not automatically be a national law. Indeed, in theory, it is possible to choose principles of international contracts such as the Unidroit Principles or the European Principles on Contract Law. In practice, however, such a choice may be unwise considering the vagueness of some of these principles and the fact that they are in a preliminary stage of maturation and have not been applied by courts or arbitral tribunals. Thus, it is difficult to know with certainty what result can be obtained by the application of such principles. In addition, by choosing these principles, the parties to the contract do not avoid the public policy of the law of the place of performance or the public policy of the place each is situated. The latter laws have a greater potential applicability where the place of performance of the contract cannot be ascertained. Indeed, the essence of an Internet transaction is the absence of location of the transaction itself unless performance involves some physical contact with a certain jurisdiction (for example, delivery of goods).
In the case of a sale of goods, the contract will be performed at the place to which the goods are delivered, which is assumed to be where the transfer of property occurs between the seller and the buyer. Thus, it may be said that the public policy of the country to which the goods must be delivered must always be taken into consideration at least for issues concerning the safety of the products and any inherent flaw in the products.
The validity of the clause of choice of law in the contract is not subject to specific formal requirements. However, the choice of law must be expressed; that is, the clause must appear in the contract itself. If not, some legal systems (e.g., Swiss Private International Law) accept that the choice results with certainty from the provisions of the contract or from the circumstances. In practical terms, it is better if the choice is made before the contract is formed. However, certain systems accept that the choice be made a posteriori but it is clearly more difficult to reach such an agreement since the parties may appear to have divergent interests when that question comes up.
When is a contract formed when concluded via Internet? To answer that question, one should apply the same rules as for any other contract formed between absentees. In our hypothetical, the contract is for the sale of goods. The answer may lie in the Vienna Convention on Contracts for the International Sale of Goods (11 April 1980). The formation of the contract occurs when there is agreement between the offeror and the offeree on the nature, quantity, and price of the goods (Articles 14 to 18). This agreement may be reached not only by an exchanged offer and acceptance in identical terms, but also if the offeree performs an act relating to the performance of the transaction itself (Article 18 (3)). If Primus sends information on its products to Secundus and/or any other specific person, Primus is the offeror and Secundus the offeree (Article 14). On the contrary, if Primus only provides information on its Web site without addressing it to specific persons, and Secundus, after visiting Primus' Web site contacts Primus, Secundus is the offeror and Primus the offeree (same Article). In any event, the Vienna Convention remains silent on the question of whether clauses, such as applicable law or choice of court, are of an essential nature, although the answer will have an impact on the formation of the contract (Article 19). Thus, an offer may be accepted and a contract formed without those clauses being agreed upon. If for Primus or Secundus it is essential that the contract not be formed unless its general conditions of sale or purchase are also agreed upon, a special mention to that effect must be included in the offer, acceptance, or counteroffer.
The proof of the clause follows the general rules of proof of contract. The Vienna Convention provides that "a contract of sale need not be concluded in or evidenced by writing and is not subject to any other requirement as to form. It may be proved by any means including witnesses" (Article 11). Also, although it is not yet "hard law," due consideration must be given to the UNCITRAL Model Law on Electronic Commerce (1996) which defines equivalents to the "written requirement."
Where there is no choice by the parties of a law applicable to the contract, the law will be defined either by the court or by the arbitral tribunal charged with the resolution of the dispute. If the parties have agreed upon arbitration (see below), the arbitral tribunal has great freedom to actually decide which rules apply. It may decide that a national law must govern the transaction, but it may also decide that it is not necessary to apply a national law since it may deem that the solution to the dispute can be found in the general principles applicable to international contracts, trade usages, or any similar body of rules it may find appropriate. It is clear that leaving the definition of applicable law to an arbitral tribunal may negate part of the certainty the parties thought they had in concluding the contract.
If the parties have not agreed to arbitration (whether or not they have elected the court which will have jurisdiction -- see below), the dispute will have to be decided by a national court. In deciding which law must be applied to the transaction, the court is bound by its own rules on conflict of laws. A Swiss court will apply the rules of the Federal Act on Private International law (18 December 1987). A German court will apply the Rome Convention on the Law Applicable to Contractual Obligations (19 June 1980), as incorporated into its private international law statute, and a French court will apply The Hague Convention on the Law Applicable to International Sales of Movable Tangible Property (15 June 1955). A United States court will not find any guidance in written law, the United States not being a Party to any International Convention on the subject, and the UCC not providing any conflict-of-law rule other than party autonomy and, in the absence of choice, its own scope of application.
Thus, a US court will follow precedents with some guidance in the Restatement of Conflict of Laws. The main difference between Continental/Conventional systems and common law systems is the flexibility of the rules. It has been argued (see Burnstein and Kromke, in Boele-Woelki and Kessedjian, Internet, Which Court decides, Which law applies?, Kluwer Law International, 1998) that Internet pleads for hard and fast rules instead of flexible ones. If the case were tried by a German court, the law to be applied to our sale of goods would be the law of the party which owes the "characteristic performance" of the transaction, i.e., in a sale of goods, the law of the seller. In our case, German law would thus be applied by a German court. If the case were to be tried in a US court, the reasoning would be more complex. Indeed, many more aspects would have to be taken into consideration, such as the language of the transaction, the currency used, the place of payment, where the goods are actually delivered, and where the goods are to be used. A US court could consider that the location of the seller and/or the buyer must not have any implication in the choice of law, since the transaction was concluded via the Internet (dematerialized transaction) (US CASE LAW ON THIS ISSUE TO BE VERIFIED).
4. In the contract, can Primus and Secundus decide which court will hear their potential dispute arising out of the contract? Can they choose arbitration? Can they provide for any alternative dispute resolution mechanism? Can they decide that the arbitral tribunal will meet virtually via Internet?
Again, as in any other contract, contracts formed via the Internet may contain a choice of court or an arbitration clause, with or without any alternative dispute resolution mechanism. The difficulty for these clauses resides in the fact that legal systems are stricter in regulating their validity than that of choice-of-law clauses. Considering this particular feature, it would be advisable for the offeree to place the choice of court or the arbitration clause prominently on one of the first pages of the Web site where the contract is displayed. Indeed, it is important that there be no doubt that the clause was agreed upon, at the time of formation of the contract. In addition, proof of acceptance of the clause by the party against which it is sought to be enforced must be easily brought. Indeed, three issues may block the validity of such clauses: (1) the clause was not formally well entered into; (2) the clause was not legally entered into, i.e., for an arbitration clause this issue is equivalent to that of arbitrability of the subject matter; and (3) the party against whom the clause must be enforced did not give its consent to the clause. At present, issues (1) and (2) are closely related and by the formal requirements for its validity, most systems also resolve the issue of exchange of consent. Indeed, by requiring that the clause be contained in a written document, or, if not, that there be a clear reference to another existing document, or that it conform to the parties' practices or trade usages, the legal systems ensure that the exchange of consent is beyond doubt.
As far as the legality of the clause is concerned, very few contracts or subject matters are outside the scope of freedom of the parties to decide on a court or on arbitration to resolve disputes which may arise from their contract. Indeed, the major concern of companies selling products or services via the Internet relates to contracts with consumers. As the law in Europe now stands, it is clear that an arbitration or choice of court clause included in such a contract may be voidable by a court as it could be deemed to be abusive. This danger is much more remote in purely commercial, arm's-length transactions.
Theoretically, there is no problem for Primus and Secundus to decide that an arbitral tribunal would decide their dispute "online" and therefore "meet virtually." At present, some projects, which are either completed or in the process of being completed, already exist, allowing for online dispute resolution. In addition, some arbitral institutions, such as the ICC, provide in their rules that a "document-only" arbitration proceeding can be conducted if both parties agree. Apart from the fact that oral proceedings may be essential for the fair resolution of the dispute, having an online proceeding will induce certain consequences, notably as far as the place of arbitration is concerned. Indeed, the place of arbitration usually carries important consequences such as the public policy to be followed, the recourses open against the award, and the default procedural rules to be applied in case the parties cannot agree on specific rules. If the arbitration is conducted online, should we decide that there is no place of arbitration? On the contrary, should we decide that the place of arbitration is at several locations (that of each arbitrator and that of the parties to the dispute)? Probably, the answer is "neither one nor the other." The place of arbitration (since we do need one) could be deemed to be the location of the Presiding or sole arbitrator. This would be a "legal fiction" until we develop rules of substitution for the place of arbitration. One major inconvenience of this solution is that the place of arbitration would be unknown until the time when the arbitral tribunal is formed. We would argue that the predictability and certainty necessary to the parties are destroyed, and thus that this is not a suitable solution. The second possible solution would be to locate the arbitration at the place of the arbitral institution. This solution would run squarely against the tradition of certain institutions to emphasize the truly international nature of their proceedings allowing for arbitration to be conducted in many different places although they are all conducted under the auspices of the same institution. In addition, this solution is unworkable for ad hoc arbitration.
Until the existing rules have been adapted or, if not possible, new rules have been developed, parties to an electronic contract would be advised to decide on a place of arbitration and provide further that this place may be agreed upon for legal purposes only if the arbitration is conducted online.
5. Secundus sold Primus' goods to end-users, via its Web site (the end-users access Secundus' Web site from all over the world). Some of Primus' goods are defective. Can end-users sue Secundus and, if so, where? Can end-users sue Primus and, if so, where? Can end-users sue Primus and Secundus in the same court and thus have a cost-effective litigation?
Two separate types of cases must be differentiated for certain aspects of the question: the first category includes cases in which the end-users are consumers; the second category covers cases where the end-users are professionals.
In most European countries, end-users (whether consumers or not) have a direct action against Primus although they have contracted only with Secundus. Thus, they can sue both Secundus and Primus in their respective courts, i.e., New York and Germany. Some countries' jurisdictional rules provide that, if there are several co-defendants, they can all be sued in the court of one of their domiciles or main establishments. However, this rule is useful for end-users who are not consumers but would oblige consumers to sue in places (either New York or Germany) which may be remote from their bases. Indeed, since we assume that Secundus' Web site is visited by clients from all over the world, many contracts will be formed with clients located in faraway countries. In addition, most consumer contracts are, by essence, contracts of limited value. Thus, it would be impractical to ask the consumer to launch a lawsuit the cost of which will far exceed the value of the suit. Can the consumer sue both companies, each at its own residence? The answer is not clear. Under Brussels and Lugano conventions, the answer could be affirmative. However, the rules of those conventions do not apply in the case of Secundus, one of the defendants domiciled outside Contracting States. In order to know whether a consumer can sue both companies in its own courts, one would have to revert to jurisdictional rules of that court. Answers could vary from one country to another. Even if the applicable jurisdictional rules permit that both companies be sued before the court of the consumer's domicile, is it fair to the defendants to oblige them to defend in as many countries as their clients are located? Is this to be considered a normal business risk of Internet sales of products?
When the end-user is a consumer, under the Brussels and Lugano conventions, it has the right to act in its own courts provided the defendant is located in one of the Contracting States and the action is brought against its co-contractant. In addition, its co-contractant must have advertised the product in the State of the consumer, and the consumer must have taken, in that state, the necessary steps for the conclusion of the contract. Obviously, this provision does not apply in Primus' case, since the product was not sold by Primus. It cannot apply in the case of Secundus since it is not domiciled in a Contracting State, unless Secundus has a branch, agency, or other establishment in one of the Contracting States and the contract was formed out of that branch, agency, or establishment. However, it must be noted that, although the Brussels and Lugano Conventions do not apply in the case of an action brought against Secundus, many countries around the world have a similar provision in their common law. Thus, Secundus could very well be sued in the country of each consumer having bought products with which it is dissatisfied.
6. Secundus is dissatisfied with Primus' products and is losing a great deal of goodwill. Primus is a company known worldwide. In order to oblige Primus to act, Secundus uploads on its Web site criticisms on Primus' acts and on its management style. Primus considers these criticisms to be defamatory. Which forum has jurisdiction to hear the case of Primus vs. Secundus? Would the answer be different if Secundus were incorporated under the laws of Switzerland?
First, Primus can always sue Secundus for the entirety of its worldwide damages in the courts of Secundus' domicile, i.e., New York.
Under German jurisdictional law, Primus could also sue Secundus in Germany, where the effects of the harmful event are felt. Although German courts are the courts of the claimant, they would probably accept to decide on the entirety of Primus' damages. On the contrary, if Secundus had its headquarters in France, the Brussels Convention, as interpreted by the European Court of Justice (see the Shevill case, 1995) would forbid the German court to hear the entirety of the claim for worldwide damages but would limit it to damages suffered only on German territory, thus obliging Primus to sue Secundus in as many territories as it suffered damages. If Secundus had its headquarters in Switzerland, the solution must be found in the Lugano Convention, but it is not certain that the Tribunal Fédéral would follow the interpretation of the ECJ as it is not compulsory for Lugano countries to follow it.
7. In order to better know its client base, Secundus created a data bank including personal data on individuals who access its Web site. The French CNIL (Commission Nationale de l'Informatique et des Libertés) happens to learn that Secundus' data bank includes information on 15,000 French nationals and that Secundus will use this data in a manner violating French law. Can the CNIL enforce French law against Secundus? If so, how?
France was among the first countries to legislate the protection of personal data and to create an independent authority, the CNIL, to ensure enforcement of the adopted legislation. The need for protective legislation emerged with the announcement by the French government, in 1974, of its intent to constitute several governmental data banks containing social security numbers of French individuals. Under the 1978 law, protection of personal data is organized around two main requirements: (i) that the implementation of automated processing of personal data requires a preliminary declaration to the CNIL; (ii) that individuals whose personal data is being sought must be informed of the consequences of such collection and of their right to access and obtain rectification of collected data pertaining to them.
If it is relatively easy to determine exactly what is protected under the 1978 law, it is more difficult to identify exactly who benefits from this legal protection and who can be sanctioned for violations, in an international setting.
Does the 1978 law protect all French citizens whose data is collected and processed anywhere in the world, or is protection conditioned upon where the processing occurs? Another possibility would be that the protection depends upon where the protected data is accessible. However, this would mean, in the case of the Internet, that French law would be universally applicable, as the Internet allows worldwide access to information.
Again, one must turn to rules of private international law. It is generally accepted that under the French system of private international law, the protection of privacy is an area of such great importance that public policy dictates that rules of law pertaining to such protection be considered as mandatory rules (lois de police). A judge would thus be required to apply such rules to situations which fall within their scope. One must then identify which criteria determine the applicability of the 1978 law in a transnational setting.
The 1978 law provides some degree of guidance in this respect. Indeed, it requires that the entity declaring the automated processing to the CNIL specify if the automated processing is intended for export between the French territory and a foreign country. The law considers to be exportation even where the processing is partly performed in France, using data previously processed outside France. Thus, it appears that the 1978 law applies each time the processing takes place, in whole or in part, in France. The question that arises is: What constitutes automated processing of personal data? It would be consistent with the broad spirit of the 1978 law to consider that the processing operation is the initial collection of data, even if such collection is performed "manually."
Hence, the 1978 law would apply if the collection of personal data was, at some point, localized on French territory. Thus, if the personal data posted on Secundus' Web site were, at some point in time, collected in France, Secundus would be under the obligation to declare its data bank to the CNIL and ensure the right of access and modification of such data to all individuals to whom the information pertains. Failure to do so would expose Secundus to criminal penalties in France.
Yet the Internet offers new means of collecting personal data which may never involve collection in France or in any country providing equivalent protection. One thinks, for example, of "Dejanews" databanks allowing collection of data disclosed through newsgroups and the establishment of personal profiles of newsgroup participants. Where the newsgroup server is located outside France, no collection of data would be deemed to have occurred on French territory.
The present situation can thus be summarized as follows: If the collection of personal data exported outside France has taken place, even partly, in France, French law applies. However, if the collection takes place solely in a foreign country, to be later exported into France, French law does not apply.
It is important to note that the 1978 law will be modified, in the near future, to comply with the European Directive of October 24, 1995, concerning the protection of individuals with respect to personal data and free circulation of data.
Given the increasing international dimension of data processing, the criterion of the physical localization of processing was abandoned by the Directive in favor of applying the law of the location of the company responsible for processing the data. The Directive will also allow for a larger flow of international data to fall within the scope of the national laws of European Community members. Indeed, even where the company in charge of processing is established outside of the European Community, the national law of a European Union member shall apply when the company responsible for the data processing has, for purposes of such processing, "recourse" to means, whether or not automated, located on the territory of such country. Obviously, the scope of such provision will depend on the interpretation of the word "recourse," but it certainly demonstrates that the European Union is indisputably seeking, through the 1995 Directive, to submit, to the greatest extent possible and admissible, the transborder flow of personal data to the provisions of the Directive.
However, in practical terms, it is most difficult to ensure application and sanction of the law to entities located in foreign countries unless they have connections and some kind of "presence" on French/European territory. This is one reason the Council of Europe decided to form a working group with a mandate to prepare guidelines for Internet users and operators to ensure the best data protection possible.
8. Finally, if the answers to questions 1 to 7 are not satisfactory, can we think about "regulating" Internet through an International Treaty? If so, what kind of Treaty should be prepared? Which Organization(s) should take the lead? Is this urgent?
As we can see from the above answers, the situation is fairly complex and transactions concluded via the Internet induce solutions which may not be entirely satisfactory. In part, this is due to the fact that conventions applicable to most questions posed were prepared before Internet existed or before it was used for commercial purposes.
Internet users would certainly benefit from a coordinated exercise under the joint auspices of major organizations charged with unification of the law (UNCITRAL, UNIDROIT, European Council, and The Hague Conference on Private International Law) so that, if a text is indeed adopted, it would be a single one, covering all aspects of the legal issues posed by an Internet transaction, and be uniformly applicable in as many countries as possible. This idea was proposed in Utrecht, in 1997, during a conference organized jointly by the Hague Conference and the Molengräff Institute of Utrecht University. Implementation of that idea can be successful only if it is supported by Internet users. Only they can determine whether the present situation needs improvement.