Internet Contracting

March 1, 1999

Chris Reed is Reader in Information Technology Law, Head of the Information Technology Law Unit and Deputy Director of the Centre for Commercial Law Studies (CCLS™), Queen Mary and Westfield College, University of London. He has published widely on many aspects of computer and telecommunications law. He is a consultant to the City of London law firm Arnheim, Tite & Lewis, Joint Chairman of the SCL and Convenor of the Society’s Legislative Working Party. He may be contacted at:

This is an edited version of a presentation to the SCL Advanced Law Conference in October 1998.

I offer here a personal view of what I consider the future development of the law on Internet contracting might look like. By ‘the law’ I do not, of course, mean English law – as the main reason for any seller’s Internet presence is to do business in multiple jurisdictions, English law is (or so the seller hopes) relevant only to a small minority of his sales. What is more relevant is the way in which the whole, transnational, body of Internet law is likely to develop. In my view the mere existence of Internet commerce will be a strong force, pressurising governments and the courts towards a common approach on many important issues.

The development of other areas of computer law, such as copyright or data privacy, appears to exhibit three consistent stages through which the law passes. The first stage might most accurately be described as ‘Anarchy’. Here the law is either non-existent (as was the case for the UK law of data privacy) or, where existing law clearly applies, it is either uncertain in its application or, more usually, ill-adapted to answer the new questions posed by the technology. The latter is well illustrated by the differing approaches to software copyright adopted in the US, Germany and Australia. This anarchy is dealt with:

  • as between the parties to a transaction, by agreeing express contractual terms;
  • eventually, through the devising of implied terms by the courts.

There is little doubt that the law relating to Internet contracts is mainly at this stage.

Stage two is where legislators recognise that the law is defective and attempt to cure its defects by passing new laws. However, each country’s legislators approach the problems from their own national perspectives, which inevitably differ. The result is conflicting national laws, which in many ways make the anarchy even worse.

The final stage is where an international consensus is reached on the proper solutions to these problems. Consensus can be achieved in a number of ways, most commonly by:

  • conventions;
  • harmonisation; or
  • convergence.

This article attempts to identify those areas where consensus may be quite easy to achieve, and ends by posing some problems about which consensus is, in the author’s opinion, much farther away.

It is already clear that a number of legal issues arise from cross-border Internet commerce which are not adequately addressed by existing laws. The most important of these are:

  • Applicable law issues. It is difficult to decide for any particular cross-border Internet contract which country’s law governs the transaction, and it is an almost impossible task to give any general advice in respect of a seller’s full range of Internet contracting activities. To some extent choice of law clauses can be used to control the uncertainties in this area, but these clauses are unlikely to be effective in all cases.
  • Responsibility for communication defects. If, during the process of contract formation, a particular message is lost, delayed or corrupted, the law needs to provide which of the negotiating parties is to bear the risk of that communications failure.
  • Information supplies in a pure form, unconnected to any physical carrier such as paper, are made possible by Internet commerce. The law needs to determine the terms which are to be implied into such a supply contract, even if only to assess any express terms against consumer protection legislation. A second issue which arises here is that perfect duplicates of the supplied information can be made by its recipient. The supplier therefore needs to identify both himself (for rights protection) and the purchaser (for licence enforcement).
  • Enforcement of obligations. A claimant will, of course, need to prove his case via computer records. In general the evidence law problems raised by computer records seem largely to have been resolved, though there is still an issue of differing rules of discovery and evidential presumptions, which will be important to purchasers because it is likely that only the supplier will have sufficient records to prove the obligation. Additionally, enforcement by the seller may be problematic because of the difficulty in proving that the specific copy of information at issue was not lawfully acquired by the defendant. A purchaser, on the other hand, will have great difficulty in enforcing a foreign supplier’s obligations if the value of the information supplied is lower than the cost of litigation.

Let us examine these last two issues a little more closely.

In the physical world most objects of commerce are classifiable either as goods or services. In either case, a developed legal system will have, over time, identified the terms which should be implied into contracts for the provision of goods or services. These implied terms are largely standard across most jurisdictions, and have often been codified in international instruments such as the UN Convention on the International Sale of Goods and the Uniform Law on International Sales.

In contrast, information products which are supplied as discrete packages and without being embodied in some physical carrier are likely to be classified as neither goods nor services. An example of this is the classification of software in the judgment of Sir Ian Glidewell LJ in ICL v St. Albans DC. The result is that there are no pre-determined implied terms. Clearly such terms will need to be developed.

What should these terms be? The existing laws on goods and services suggest that, at a minimum, implied terms should cover delivery, ownership and use rights, and description and quality. However, in the case of information products, the content of these terms is likely to be very different from their physical world analogues.

Implied terms as to delivery of goods or performance of services tend to be simple; delivery or performance within a reasonable time, unless a specific time is agreed. For Internet-supplied information products, however, this is unlikely to be specific enough. I suggest that the likely content of such a term might be:

  • if the information is transmitted by the supplier, eg via an http download or an e-mail attachment, it should be transmitted within [a reasonable time/X hours] of payment. Internet commerce in such products will require payment before delivery, and thus the completion of the payment process should trigger the time period;
  • if the information is to be collected by the purchaser, eg by providing him with an ftp password or creating a customised Web site with password-protected http download, it should be made available for collection within [a reasonable time/X hours] of payment (for the same reasons) and remain available for collection for a reasonable time.

Additional implied terms will also be necessary to deal with communications failures in the file transfer process:

  • that the seller will provide for redelivery if delivery fails; and
  • that the seller will make refund if delivery is impossible, and in a similar form to the payment method used to purchase the information.

Ownership and use rights are at first sight simpler. By analogy to the provisions for goods, the purchaser should receive the information free of third party intellectual property (and other) rights which prevent or limit the purchaser’s use of the information. The difficulty here is to determine what uses by the purchaser should be protected; he will not be the owner but a licensee of the information and should therefore receive limited use protection. Should that protection cover common uses of the information, extend to expected uses, or even to the purchaser’s subjectively intended use? In many cases this question will be answered by express licence terms, and in others the courts may well decide to determine the scope of use rights by constructing an implied licence. The Australian case of Trumpet Software Pty Ltd v OzEmail Pty Ltd suggests that such an implied licence might allow those uses which do not adversely affect the ‘owner’s’ ability to exploit the information product commercially.

Quality terms will also be needed, and will be important for liability purposes as more and more reliance is placed on commercially purchased information products. These terms are likely to be based on conformity with description and fitness for purpose, but the latter will prove problematical. For supplies of goods, the courts have been able to identify a limited range of ‘common’ purposes for which they must be fit, but digital information is so versatile that this may prove difficult for information products. Purchasers of physical-world services tend to specify their precise purpose, but this is unlikely in a mass-market for online information. I can offer no real assistance in predicting how this issue will be resolved.

Enforcement of the seller’s rights in his information products will depend very much on encryption technology. A successful infringement action will require proof of two important matters:

  • that the defendant is not a licensee or otherwise a lawful acquirer of the information. Because many Internet dealings in information will be with unknown purchasers, the seller will need records which adequately identify lawful acquirers. The most effective way of achieving this will probably be to embed the purchaser’s digital signature in the information product. To be effective as evidence, such signatures must have been produced via a proper digital signature infrastructure, and current legislative initiatives are moving towards a partial consensus on the conditions under which such an infrastructure should operate. A related issue, already addressed in legislation such as the Software Directive, is the extent to which the purchaser should be able to transfer the information product to a third party;
  • that the information originated from the seller and not some third party or via independent creation. This will be particularly important for micro-products such as clip art or non-creative information, eg a share price. Encryption-based technologies such as fingerprinting and watermarking will be useful here, but will need the benefit of evidential presumptions.

From the consumer purchaser’s perspective, a number of legal issues will need to be resolved. Perhaps the most obvious is the need for a consensus on whether a seller can contract out of the consumer’s domestic consumer protection laws. This need arises because it is by no means certain that the consumer is not ‘visiting’ the seller’s jurisdiction when he makes his purchases, and in the physical world such a consumer would not expect to benefit from his domestic law.

A more practical issue, which may nonetheless turn out to be of real importance for consumer confidence, is that of enforcing the consumer’s rights against the seller. Most Internet transactions, and particularly those for information products, are likely to be low value. Consumers are unlikely to be prepared to bring legal proceedings in a foreign jurisdiction to recover small sums. The most likely solution will be for third parties, such as consumers’ organisations, to be able to enforce rights as representatives. This can be achieved through voluntary schemes, which might prove effective if they also achieve marketing benefits by enhancing consumer confidence, or through legislation as has already been suggested in the Distance Selling Directive.

To conclude, I mention a number of further issues on which there is, as yet, no sign of an emerging consensus. The following is by no means intended to be an exhaustive list:

  • Contracting through software agents, sent out to roam the Internet and negotiate bargains with sellers’ software poses problems. Contracting requires, in theory, a meeting of minds. Common law jurisdictions have many tools with which to find such ‘contracts’ valid, estoppel being perhaps the most obvious. Those civil law jurisdictions which lay a greater emphasis on the mental element in contracting may have difficulty in finding such agreements to be binding and enforceable.
  • Formalities such as writing and signature abound in national laws. All countries will need to re-visit these requirements and amend them to enable Internet commerce to take place.
  • Joint liability of intermediaries is a concept which already exists in the law relating to credit cards and has been proposed as a consumer protection measure for Internet transactions. The theoretical basis for imposing such liability will need to be developed, so as to discriminate between different classes of intermediary, before joint liability for some classes can safely be adopted as a general measure.
  • Finally, the conflict of laws problem will need to be simplified. Although private international law offers all the tools which are necessary to resolve the law and forum questions, these tools can only work on a case-by-case basis. This will prove too complicated and expensive for low-value consumer transactions. The simplest solution is to prohibit choice of law, at least in consumer transactions, and provide that the purchaser’s law applies. However, this approach makes the assessment of the seller’s legal position extremely complicated. It may be that this is an issue on which consensus will not be reached for many years, if ever.