Exploring the Frontiers of Cyberspace - the Challenges of Information Technology beyond the Year 2000
"Cyberspace" is a term coined by science fiction writer William Gibson. He used the expression to describe a "place" where data is stored and transformed. Characters in Gibson novels find nothing out of the ordinary in exploring Cyberspace in order to look for changes in the "Data Matrix" which will signal significant changes in the world order. Cyberspace is the "place" where ordinary telephone conversations "happen", where voice mail and e-mail messages are stored and exchanged and where computer generated graphics images live.
Modern legal writers have extended the term to include all forms of computer mediated communications and interactions. We are speaking here of communication forms as familiar and ordinary as everyday telephone conversations to forms as esoteric and extraordinary as universal optic fibre connections or personal bi-lateral or multi-lateral tele-video communications.
Last week the local Sydney newspaper, the Sydney Morning Herald, contained a report of a meeting which was debating the need to censor video games. It seems that some games available now for personal computers contain material which is thought by some to be pornographic. The meeting voted by a bare majority in favour of censorship.
This should not surprise us since nearly every form of communication technology has been received with suspicion and hostility by some sections of the society. In the early part of the century a German psychiatrist claimed that the telephone was responsible for driving people permanently insane. In more recent times, the use of special numbers which provide special services has caused concern among members of the society who are responsible for being concerned about such matters. This form of communication, which Professor Eli Noam has called "tele-publishing", has been restricted in one way or another in many countries.
The introduction of mass television broadcasting sparked a flurry of research by sociologists and psychologists on the effect of the new technology. Findings were, and are, nearly always adverse to the medium. Again drawing from recent experience, another newspaper report was of a researcher who found a correlation between suicide in young people and the number of hours of television watched. The research, and the findings, reflect early research following the introduction of radio broadcasting which noted that "parents have become aware of a puzzling change in the behaviour of their children." The findings generally fail to mention that all changes in the behaviour of children are puzzling to parents.
Most of you will be aware of the concerns expressed about the effect of computers on children and others, so I will not continue to elaborate upon this theme. The point is that Cyberspace has presented problems to us in the past. Some of the problems were real enough, some of them existed only in the minds of a few whose interests were somehow threatened. In each case we have either solved the problem or adjusted our behaviour so as to avoid it. In my view, studying the way in which we have solved past Cyberspace will provide a good guide to solving future ones. I would like to turn now to a discussion of those future problems.
The seminar rooms of the world are littered with the bodies of those who tried to make accurate predictions of the shape of computer technology over the next ten years. In general we can say this: hardware development will proceed faster than we predict and software development will proceed more slowly. This is unfortunate because, as we will see, software is the more important.
We can safely predict that wide-band communications channels will continue to be more widely available to a greater number of people in the world, probably through the physical medium of fibre optic. But, in a development which contradicts my above comment about the speed of software development, compression software is now allowing us to transmit remarkable amounts of information, including video images, along ordinary telephone lines. This is a very important development since the telephone infrastructure is already in place in most of the world.
We can also safely predict a substantial increase in the number of bi-lateral and multi-lateral networks. By a "bi-lateral" network I mean one where most messages are transmitted between two people at any one time. The models for this are the telephone and private mail messages. By a "multi-lateral" network I mean one in which messages are generally intended for a wider audience. This could be a bulletin-board type of network where messages are "posted" for later reading by members of the network or one in which "real-time" multiple party communications occur such as a conference telephone hookup.
For the immediate future most of these networks will probably be confined to text and voice transmission, but we should also expect to see bi-lateral and multi-lateral video networks emerge in the time frame with which we are concerned.
Networks are not be limited by geographical or political boundaries, but this is not to say that they will necessarily be publicly available. Some may be widely available to a large class of people: the academic "Internet" is an example of such a network. Or there may be special purpose networks whose clientele is limited and which are intended to serve a special purpose: the banks' SWIFT system is an example of this type.
An individual who is using a major network soon discovers a serious problem. There is too much information. This is true whether the network is a general purpose network or a specialised one. The natural tendency is for information to expand to fill the facilities available to transmit it. When it cost real money to send messages by mail, I would receive perhaps five to ten items a day related to my work. On the Internet it is not an unusual day when I receive fifty or more messages.
There is a natural reason for this. It is not difficult to add information to a network, particularly if you are not overly concerned with the quality of the information being added. Television provides an excellent example of this, but it happens in all networks.
The personal dimension of Cyberspace then is this: how does one deal with the amount of information received? There are not too many strategies available to us to solve this problem. One is the simple expedient of allocating more of our time to perusing the incoming messages. This seems to be the strategy adopted by those consumers who have cable television services: one report shows that the television is on for more than eight hours per day in such households.
Another strategy is to package the information so that the user can quickly assign priorities. For example, in the Internet electronic mail system each message has a "subject line" which should give an indication as to the contents of the message. This allows the user to read only those messages which are of most immediate interest, postponing, perhaps indefinitely, the reading of the less interesting material. Less ennobling, each message also indicates the identity of the sender, so that messages from the Head of Department may be deferred.
The problem with this strategy is that it depends upon the good will of the sender to identify the contents of the message in a way that is meaningful to the recipient. Our experience with television should warn us that the sender of information cannot always be trusted to provide an impartial and accurate account of the contents of the transmitted message.
What is needed is a "smart" screening system that assigns priority to incoming information based on the wishes of the user of the system. "Junk mail" could be filtered out entirely and priorities assigned to the remainder.
Unfortunately, a smart screening system is closely related to the general problem of language understanding. This is a problem which has proved to be remarkably intractable for computer science, and although one reads occasional newspaper reports of amazing new developments the reality is that a machine which "understands" human language is still not within our grasp and is unlikely to be so by the year 2000.
Cyberspace knows no geographical boundaries, and this in itself is a cause for political concern. For the first time in history small interest groups can form with membership which is widely scattered. It is much easier to achieve the "critical mass" which is necessary for a group to be self-sustaining. It is easy to imagine that the formation of such groups will be of concern to governments.
A more immediate concern, however, is the general framework of ownership and control of the networks. Professor Eli Noam has observed that networks become political entities. They acquire quasi-governmental powers as they take on the tasks of mediating the members' conflicting interests, creating their own rules of admission and exclusion and even establish their own de-facto taxation mechanisms.
A recent example from the Unites States illustrates the problems that can and do arise. The Prodigy network has more than 1.75 million subscribers. It runs some 400 bulletin boards which display more than 80,000 messages each day. There is an initial joining fee and a monthly subscription fee. Subscribers must agree to a standard form contract.
In late 1989, a bulletin board, or "forum" conducted by Prodigy called "Health Spa", became the scene of a debate between homosexuals and fundamentalist Christians. Fundamentalists posted messages attacking homosexual lifestyles as sinful and were in turn bitterly condemned as hypocritical. Prodigy closed the bulletin board that December, citing a lack of user interest.
In October 1990, Prodigy announced new, per-messages charges for email. A group of users posted messages on an open bulletin board protesting and calling for a boycott of Prodigy's advertisers. Prodigy then announced that such messages were no longer permitted and refused to post them on bulletin boards. Protesters then began to use private email and also posted email messages to the advertisers. Advertisers revealed these messages to Prodigy who in turn terminated the subscriptions of the protesters and refused to reinstate them. Prodigy has also announced other restrictive rules concerning the use of the email facility.
In a third incident a racist message had been circulated via the email system. It had been submitted for a bulletin board posting but rejected. Prodigy argued that it could do nothing about the contents of private email because of federal law.
The focus for us here today is to note the problem. However you feel about the outcome of the particular Prodigy dispute, it provides a clear example of the kind of problems which we will face. The Prodigy dispute is (perhaps) easy to dismiss since it involves no direct threat to personal or national livelihood, but the same powers of exclusion could have catastrophic effects on business if exercised in a commercial network.
As a hypothetical example consider a trading (EDI) network which wished to impose rules on the types of goods which it would allow to be traded. Since EDI messages must conform to a limited format it is easy to build an automatic scanner which would alert the owners of the network of any breaches of their imposed rules. The non-geographic nature of networks means that the owners and controllers of them may not be sufficiently responsive to national concerns or sufficiently vulnerable to national regulation. We can also expect the emergence of "data havens" which will play a role similar to that played by certain countries in banking and finance.
All of this is background to the major theme of this conference, that of electronic commerce. What are the problems likely to be faced by someone who wishes to do business in Cyberspace? Since this is the main theme of later speakers, I will merely provide a few examples to set the stage for their deeper discussion.
Business in Cyberspace will be faced with all of the problems and opportunities outlined above. In addition, there are some problems which are particularly relevant to business and yet others which are particularly relevant to the law and legal change. These are offered by way of example only, not as an exhaustive catalogue.
One of the attractions of business in Cyberspace is the reduction in the amount of paper which needs to be processed. Particularly in international trade this can amount to a significant amount of the overall value of the transaction. An American study some years ago estimated that paper processing could amount to as much as 20% of the FOB value of an international sale of goods.
One example will illustrate the types of problems which occur when paper is replaced by electronic signals. The bill of lading serves as a document of title and evidence of the transferable contract of carriage. The detailed contents of the document are legally important, and it is vital that illicit changes should not go undetected. But it is equally important that there is only one original since possession of the document confers rights.
Paper is not a perfect medium for carrying the messages of the bill of lading. Alterations of documents are not unknown, and forgery is a possibility which every legal system must contemplate. However, it is not easy to make an undetectible alteration of paper, and most forgeries are not of the skilful variety found in the best detective novels. Consequently, we can be reasonable certain of the integrity of the paper document, that is, that the message has not been surreptitiously changed, and of the identity of the parties to the document.
Paper also has the admirable quality of being only in one place at any given time. It makes sense to speak of possession of the document whereas speaking of possession of a message sounds artificial. It is this quality of paper which allows possession to assume legal significance.
If we are to devise commercial practices which allow electronic documents to replace paper bills of lading then we must solve three problems: the integrity problem, the identification problem and the uniqueness problem for electronic messages. There are two general approaches to solving these problems.
The first approach, the organisational approach, is to create a trusted depository. The basic idea is that the parties to a transaction would all communicate via the trusted depository. The depository would be responsible to guarantee the identity of the parties with which it is communicating. The integrity of the message is assured by the depository. The uniqueness problem is solved by asking the depository who "possesses" the document at any particular time.
The trusted depository scheme has been implemented in several different contexts. The Intertanko scheme for bills of lading is a depository scheme, primarily proposed to solve the problem of stale bills. The Austraclear scheme in Australia is a depository scheme which allows the electronic trading of finance bills, although the paper bill has not been entirely abolished.
The second general approach to the problem is a technical solution. Is it possible that the message itself can carry with it evidence that it has not been altered? Is it possible to devise schemes whereby a message can convincingly say, in effect, "I am what I purport to be"? The somewhat surprising answer is in the affirmative by use of what is known as public key cryptography.
The essential feature of public key cryptography is that the message is encoded by the sender using a secret key, but is decoded by a publicly available key. A message is accompanied by "plain text" which advises the identity of the sender. If the use of that sender's public key can decode the message, then it must have been sent by that person. Equally, the contents cannot have been changed since the sender is the only person who knows how to encode for decoding with the public key. Various other combinations of encoding/decoding using the public and private keys allow for other orderly document modifications such as electronic indorsements.
Neither of these solutions should be though of as "best". Each may be optimal in certain circumstances, and the benefits and costs of each must be carefully considered in the commercial context before further decisions can be made.
In theory, business in Cyberspace can be run from any geographical location. This has significant consequences for countries where many local industries are subsidiaries of international corporations. Specifically, local autonomy of these subsidiaries will certainly be diminished. Even low level every-day decisions affecting the subsidiary can be made at remote locations by individuals who have little understanding of the local sensitivities.
The centralisation problem is not a new one, but the change in degree may be so substantial as to be a change in kind. Solutions to the problem must clearly be political ones, and we may expect that they will be slow in coming.
The legal problems of EDI are the topic of a later session and I will not discuss them further here. However, it would be wrong to think that the legal problems raised by information technology are confined to EDI.
For example, the law of copyright has undergone significant strain and adjustment over the last twenty years in order to adapt to the problems raised by information technology. We can expect further changes as the courts of every country struggle to identify the boundaries of the idea/expression dichotomy in computer programs and databases.
Again, most countries have had to reconsider their laws of privacy, if for no other reason than to meet international standards for transborder data flows. Countries which fail to meet minimum standards of privacy law will find that their commercial enterprises are locked out of markets which call for the free transmission of personal data.
Artificial intelligence is that branch of computer science which is concerned with developing machines and programs which perform tasks that we generally think of as reserved for human thought. As an example of the current legal capability of computers, Appendix A contains a legal opinion written by a computer.
AI will pose several interesting opportunities and problems. For example, it is already possible to have computers which will order stock automatically when reserves fall below a certain level. The retailer's computer may directly contact the supplier's computer and issue the order instruction. The supplier's computer will arrange for the dispatch of the goods and issue the appropriate invoices. It is in all respects an ordinary contract for the sale of goods, but what happens when there is a dispute as to the quality of the goods or some other term of the contract? What happens if there is a dispute which requires us to decide when or if a contract has been formed? What "representations" have been made by one computer to another? Has there been "reliance" on these representations?
In another area, one of the promises of AI is to provide more consistent administrative decisions. Licences may be granted or refused, certificates issued or refused and entry into a country permitted or prohibited in the first instance by a computer. A machine can make these decisions rapidly and fairly based on policy determined criteria. What kind of appeal mechanisms must we implement to protect against unjust machine-made decisions?
The conclusions are short. Cyberspace presents us with new challenges in areas of social policy, commerce and law. Every challenge presents both an opportunity and a threat, but this is not new in the history of human technological development. The overall challenge is to make maximum use of the opportunities and to minimise the threats. Conferences such as this which allow us to focus on the issues and discuss possible responses contribute directly to meeting these challenges.