Testimony before the

House Committee on Commerce


on the subject of

The Future of Electronic Commerce


 Rayburn House Office Building

U.S. House of Representatives

Washington, D.C. 20515



Dr. Robert E. Kahn, President and CEO

Corporation for National Research Initiatives

1895 Preston White Drive, Suite 100

Reston, Virginia 20191


April 30, 1998



I am pleased to be here at the invitation of the House Committee on Commerce to testify on the future of electronic commerce.

My name is Robert E. Kahn and I am President and CEO of the Corporation for National Research Initiatives (CNRI), a not-for-profit scientific research organization established in 1986 and located in Reston, Virginia. Prior to founding CNRI, I was employed by the Defense Advanced Research Projects Agency (DARPA) from 1972-1985 where the last position I held was director of the Information Processing Techniques Office. Before that, I was on the technical staff of Bolt Beranek and Newman (now a part of GTE), the faculty of MIT and the technical staff of Bell Laboratories. At BBN, I was responsible for the system design of the ARPANET, the pioneering computer network that led directly to the Internet. At DARPA, I was involved in the development of advanced information technology where we funded many of the ideas embodied in the basic technology used in the Internet today.

CNRI is developing infrastructure capabilities to support electronic commerce, and is active in research and other activities that relate to the Internet. This technology is being used by a number of groups in the private and public sectors, including the Library of Congress. CNRI also facilitates several cooperative activities in support of Internet infrastructure operations, standards and applications. These include an Internet service providers group known as "IOPS.org," a multi-industry group known as The Cross-Industry Working Team (XIWT), the Internet Engineering Task Force (IETF) Secretariat, and the D-Lib on-line journal for the digital library research community.

Many studies of electronic commerce have been conducted by others including the recent Department of Commerce report entitled "The Emerging Digital Economy" (at http://www.ecommerce.gov/emerging.htm) and the Cross-Industry Working Team white paper entitled "Electronic Commerce in the NII" (available under White Papers & Reports at http://www.xiwt.org). Multiple scenarios of future electronic commerce have been developed, and many of the fundamental issues pertaining to electronic commerce have already been identified. No attempt is made here to portray those studies or to recreate their findings. Nor have I attempted here to define the concept of electronic commerce.

As requested by the Committee, I focus here on some personal views relating to electronic commerce, and identify several areas where the U.S. Government, acting on its own or in concert with industry, can help to facilitate its development. I take it as given that the private sector has the lead role in creating goods and services, and in making investments and reaping rewards for competing successfully in the marketplace. But I believe that there are very legitimate, even crucial public sector roles as well. In supplementary remarks at the end of the testimony, I elaborate a bit on the nature of digital information in the future, and the management of access to information on the Internet.

Let me offer two general observations. First, I believe the rapid growth of electronic commerce in the past year is largely a result of having started on a small base. An important factor, however, is the expectation that critical public policy decisions will be made to insure levels of trust and reliability in the infrastructure, which warrant future growth. The Government needs to accelerate its efforts to promote overall system dependability, to make it trustworthy, and to keep it open to innovation. If this fails to occur, electronic commerce may not continue to grow as fast as has been widely predicted.

Second, it is critical that the issues involving commercial applications and commercial practices be considered separately from challenges of developing and deploying infrastructure. They are interrelated, but separable, and both need to evolve properly. The Government has a role and responsibility in infrastructure matters that go beyond private sector capabilities to provide goods and services.

My specific observations discussed below deal with: 1) maintaining a leadership role for the US in electronic commerce; 2) insuring a reliable underlying infrastructure for electronic commerce and a level playing field for competition; and 3) helping to make digital information a "first class citizen" in the network environment.


1. The U.S. needs to maintain a leadership role in electronic commerce.

Our leadership in electronic commerce has many sources. One important factor is a steady stream of advanced technical knowledge and trained people who can exploit it. Significant parts of today’s computing and information technology base result from federal research investments of the past few decades. The nation reaps great rewards from these investments, in the form of new products and services, increased productivity, and wealth creation.

We need to continually replenish the supply of innovative ideas and human resources upon which industry relies. To accomplish this most effectively, the federal government should make it a high priority to fund advanced research in information technologies through the well-tested processes of its research agencies. This research should include the large-scale testbeds and processes that promote advanced infrastructure deployment. That is how the Internet came to be. Industry, the record shows, will work with the government in this endeavor.

Beyond advanced research, industry alone should develop and deploy commercial infrastructure innovations and applications. The marketplace generates growth. Where barriers and obstacles exist to continued competitive growth, government’s help may be required to remove them.

2. A means of insuring a reliable underlying infrastructure for electronic commerce, and a level playing field for competition, is necessary.

The private sector must lead in taking risks in the marketplace. Government has a responsibility to insure a reliable and dependable infrastructure, and to see that the needed measures to insure trust and efficiency are in place. Government needs to prevent barriers and obstacles to the competitive growth of electronic commerce.

This is complicated by both domestic and international commerce, with jurisdiction shared by multiple authorities, whether they are different nations or different states. Experience in the communications arena may be of use in deliberations on how best to approach these issues. However, any such measures--technical, legal or otherwise--should be considered with great caution in light of important First Amendment and privacy considerations.

The most fundamental aspect of trust is in insuring that the infrastructure will continue to operate, or can be rapidly reconstituted, even in the face of failure on the part of the private sector. Mechanisms must be in place to recover from widespread network attacks of various kinds, ranging across viruses, malicious hacking, accidental and purposeful attacks. Like national security systems, these may be too complex or otherwise uneconomic for adequate implementation by the private sector.

3. Digital information needs to be a first class citizen in the networked environment.

The fundamental attribute of digital information is that it is processable data, or even active elements enabling new forms of electronic commerce and creativity. Digital information can be similar to or quite different from other forms of information that we are used to dealing with outside the networked environment. When it is like what we are used to, it should be possible to use existing law and practice to deal with it for commerce. But where it is not like what we are used to, we may need to deal with it in new ways. It is at this frontier that the Government needs to be active if it wishes to facilitate new forms of electronic commerce, new opportunities for wealth creation and social benefit. Sometimes this will be straightforward to achieve, other times not.

Many existing practices and technologies can be carried over to new electronic commerce applications in the future, but new ones are needed as well. Deciding how best to proceed will require the active participation of government. Some important areas for deliberation are as follows:

The Government should participate with the private sector in making digital information a "first class citizen" in the networked environment. This should involve at least three important attributes: unique persistent identifiers, additional data about how to manifest this information at future times, and an efficient way to determine terms and conditions for access.

Similarly, the need exists across the board for the Government to better manage its own information, some of which will be an important building block for electronic commerce. In some cases, certain information is available on the Internet today, but other information is of such a nature that it cannot be made available in open source form. Classification and privacy concerns, for example, may preclude such access. Methods for archiving the material so that it can be appropriately cataloged indexed and accessed will be essential. This is evident from such recent rulings as the requirement that government email must be archived, but the needs will likely go far beyond that.

An open system with defined interfaces, protocols, methodologies and objects will best serve this purpose. It will enable a whole new dimension of electronic commerce to develop in the provision of platforms and software to meet the needs of the open architecture framework. One such framework, called the "Digital Object Infrastructure," is under development by CNRI. This is an open architecture, non-proprietary long-term approach to this problem, which I believe would be an excellent candidate for the Government to consider in its deliberations about how best to manage structured information in the network environment.

Our starting point is that digital information may need to be long-lived and may have rights, interests, value or constraints associated with it that must be observed. Technology components within the framework include 1) a system for unique, persistent identification of the structured information we call "digital objects"; 2) repositories for storage and access to digital objects, and; 3) registries for maintenance of authority information such as might be used to authenticate digital objects in various ways. The XIWT white paper entitled "Managing Access to Digital Information: An Approach Based on Digital Objects and Stated Operations" (available at www.xiwt.org) provides useful background material on this framework. Other approaches may be under development elsewhere and may be worth examining as well.


Electronic commerce is already a major component of the economy and should grow significantly in the future. Government has a critical role to play in insuring that the electronic marketplace has a level playing field, operates efficiently, and can be trusted by its users. Of particular importance are the support of advanced research as an investment in the future and the elevation of structured information as first class citizens in the network environment. Government should take the necessary steps to insure that electronic commerce continues to grow rapidly in the future.


Current network applications which enable the growth of electronic commerce today derives primarily from research and development investments made during the 1970s and 1980s. Continued research and development in networking and network-based computing is required to leverage the enormous potential of the newly emerging computing and communications technologies. A steady stream of innovations, the normal byproduct of advanced research, will be needed for the U.S. to maintain its leadership position.

While industrial investments have grown significantly in this area over the past two decades, they will not necessarily guarantee the needed innovative processes. In particular, industry investments (vs. government supported research) tend to have the following characteristics: short-term vs. longer-term, proprietary vs. open, opportunistic and market-driven vs. advanced or knowledge-based, and product-oriented vs. cultivation and development of individual talent and ideas.

A critical dimension of the new electronic commerce landscape is the presence of content or other material in which intellectual property rights and interests may be present. New approaches will be needed to manage this type of material and to support wholly new creations or innovations for which there is no counterpart outside the computer network environment. Of particular importance will be the need to identify such material (uniquely and perhaps over long periods of time), to locate such material and to manage access to such material. I expect one example of a wholly new creation to be "malleable content" which is made available on the Internet for the express purpose of having others modify it as they see fit. New economic models may be needed for such activities and new types of infrastructure may be required to facilitate those economic models.

Today, the roles of government and the private sector have been worked out over many centuries and mostly work quite well. However, certain issues such as how to deal with security (including encryption), how to protect individual and organizational privacy, and how to limit exposure to objectionable content present new challenges. Personally, I believe we will have to accommodate a world that increasingly uses strong encryption technology, a world in which objectionable material lurks around every cyber-corner and a world in which continued vigilance is needed to protect privacy as if it were synonymous with individual freedom. Fraud, viruses, and other unwelcome activities will abound as they do in the physical world and we will have to learn to deal with them. However, disputes and dispute resolution mechanisms will also be commonplace in the world of electronic commerce as they are in the physical world.

Over the past three decades, the technology of computing has been transformed from the large main-frames characteristic of the early computers, to increasingly smaller machines now dominated by state-of-the-art microcomputers that run many orders of magnitude faster than their earlier counterparts. Computer networking technology emerged in the late 1960s with the ARPANET, and has led to the Internet and increasingly higher communication speeds. Local area networks and associated servers are now increasingly the norm at offices around the country and may soon exist in larger numbers at homes as the normalized cost of both computing and communications continues to drop.

Recent trends in optical networking, and the recently established Next Generation Internet (NGI) program, offer the possibility of ubiquitous communication speeds in the range of billions of bits per second or more to individual workstations around the country. With the availability of such high-speed networks, new technological computing options also become possible. In particular, it becomes possible to contemplate computers that are able to quickly download virtually all their software over the net. While this approach makes it possible for an individual user to customize his or her computer, the more likely outcome is that software bundlers will offer pre-configured complete packages of software and that users will come to trust certain bundlers to provide their software needs. This will represent a new form of marketplace in which new tools and techniques will be needed to insure its effective operation.

As a result, I expect that we shall see levels of electronic commerce, which will include end user applications, a marketplace for information technology itself, and the underlying infrastructure, which enables them. For the most part, existing laws and legal frameworks may be appropriate to deal with item 1 without much reinterpretation. However, we should expect to see a combination of old non-network applications ported to the network for convenience or efficiency as well as new network applications for which no prior counterpart may have existed. For items 2 and 3, it may be necessary to adapt existing laws and legal frameworks to the new market mechanisms and infrastructure capabilities.

The nature of information in digital form is sufficiently different from the print-on-paper world that it is merits examination as a driver for new forms of electronic commerce on its own. In addition to its use for traditional content purposes, new forms of creativity will be unleashed by the ability to manipulate information in digital form, as well as digital information that has inherent capability for exhibiting behavior such as distributed task execution.

Applications such as banking, accounting and medicine all require the ability to handle digital objects with appropriate levels of access control. Even official records of all kinds need to be structured, identified and made available so that they can be appropriately accessed in the network environment. A common framework to achieve this is the most effective course of action, provided that the framework focuses only on the basic infrastructure needs that are common to all such information access applications. Critical to this framework will be means of uniquely and persistently identifying such information and means for structuring it so that the means for manifesting the information can be obtained (essentially automatically by computer) in the future. With such an approach, it would be relatively effortless to move large information bases from one platform and software base to another as the technology options change.

A particularly interesting area to consider is the use of networks to facilitate engineering design and fabrication of components or even whole assemblies. An important collateral area will be the ability to collaborate with teams of designers and engineers that may be distributed at different locations. The results of such activities will have to be shared, stored, possibly reengineered or augmented at later times, compared with previous designs, or simply submitted or filed away as official records or company history. The need to manage information in this form will become increasingly apparent. CNRI is pursuing such information management systems for designers of micro-electronic-mechanical systems (MEMS).

MEMS devices are micro-electro-mechanical systems that can perform a wide range of physical functions in miniature. Examples are miniature actuators, motors, movable mirrors, with a broad range of applications including sensors for activating automobile airbags, motion sensors for automated control systems, computer displays, and medical devices. The use of the network to facilitate state-of-the-art MEMS design will enable a whole new generation of creativity and economic activity on the Internet. These designs could be made uniquely identifiable and stored in a way that others could understand the terms and conditions (if any) that would enable others to make use of them.

In summary, the possibilities for electronic commerce are so large that we can only begin to grasp the potential yet unrealized. Much effort will be placed on converting existing applications for the network environment to take advantage of its broad reach and efficiency in the marketplace. The biggest payoff for society will likely come from the new possibilities that could not be realized effectively in any other form without the network.


[ home | about CNRI | programs | news | publications
special interest topics | site map ]