In the matter of )
)
The Internet )
and the )
Management of )
Objectionable Materials )
My purpose in this hearing is to provide the Committee with current basic information in five areas:
This openness and dynamism free from the fetters of governmental regulation has been enormously successful by any measure. It has markedly enhanced science worldwide - with estimates that 70% of all the scientists who have ever lived are now accessible via the Internet. As a marketplace, internet products and services now produce revenues approaching US$6 billion.
The robust open market and institutional freedom of the Internet have incented its thousands of developers not only to devise new tools, but also to fix problems - rapidly and effectively. New efforts are now underway to fix the current problem of an Internet being used by a comparative handful out of the tens of millions of users - some with malicious intent - to disseminate materials that others may find objectionable, particularly for children. The solutions are in the form of elegant new reader selectivity techniques and tools that can allow individuals or custodians to preselect available information based on labels or third party ratings.
However, like many tools, there is a dark side with long-term implications. The actions of a comparative few abusers of the Internet and the surrounding hype have the potential of diminishing the long-term global openness of the medium, as regimes and institutions define "objectionable" in narrow political, cultural, or religious terms. Scientific research, worldwide open society and international human rights are potentially losers in the process. This should equally concern Congress.
The Internet consists of approximately 7 million computers seamlessly integrated using a common technology via 60,000 largely private networks. A little less than half of those computers and networks exist outside the USA - spread among 100 different countries. There are currently several hundred commercial firms worldwide that specialize in providing interconnection into the global internet. This agglomeration of millions of computers is a direct descendent from the original Internet assembled by the U.S. DOD in the late 70's and early 80's, and allows any one computer to have immediate and direct (but controlled) access to any other connected computer, its information, and certain processes. What those computers and their users can accomplish is limited only by the kind of computer applications on the machines.
The Internet's key features are very significant. It is highly distributed throughout the world, and the computers and networks are overwhelmingly in hands of literally millions of individuals, companies, and institutions. It was fostered by decades of government policies - domestically and internationally - that left computer networking wholly to a highly competitive marketplace and individual initiative, and is referred to as "bottom- up infrastructure." It operates over virtually every kind of underlying means, including: local networks (LANs), telephone lines, ISDN, CATV, wireless, cellular, private and common carrier fiber, satellite and submarine cable circuits. A common joke is that "the Internet runs over everything except wet string." The Internet operates as a highly distributed intelligent network that can auto- matically learn and adapt to dynamically route traffic over myriad alternative routes. A message or even pieces of a message may go different paths to an end destination at any time. Enterprise Internets, which use the same technology, also exist on a large scale in many commercial and governmental institutions. There are hundreds of thousands of such networks. However, they are not generally publicly accessible and thus probably not relevant to the Committee's hearing, but indicative of the complexity of the environment. On-Line Services are stand-alone commercial offerings of providers that allow their customers dialup access to their computer systems and a variety of fixed services on those computers. These exist separately from the Internet - although most of them can now exchange at least Email to the Internet via gateways. Bulletin Board Services are similar to On- Line Services, but generally operate at the community level on a relatively small scale, and most are not presently connected to the Internet, although this is changing rapidly as most new BBS software has Internet access built-in.
Other computer networks exist on a large scale, and have gateways to the Internet - largely for just EMail. These include, for example, UUCP networks, FidoNets, Bitnets, X.400 networks. Many of these are still used extensively in developing countries, where they were built using available computers and dialup telephone lines. However, low-cost Internet technology has recently become available and is now being implemented in many poor countries and regions. X.400 is principally a telephone company commercial Email service.
Cyberspace, by contrast, consists of all of the above, arguably including even stand-alone computer systems that may access materials physically transported by CD-ROM, diskettes, and tape.
Perhaps foremost, it would be exceptionally difficult for government at any level to actually govern operation and use, or dictate conduct that applies to the Internet as a medium. It is a mas- sively shared, constantly evolving, global system for which the only model of comparable complexity may be the global economy. Furthermore, a 30 years legacy of basic regulatory policy at national, regional and international levels have proceeded on the premise that government should forebear from regulating the computer networking business. This includes the FCC's Computer I, II, III trilogy, the European Union's Green Paper, and the World Trade Organization GATS treaty. Indeed, the growth and dynamics of computer networking environment have confirmed the wisdom of those policies.
Furthermore, any exercise of legal jurisdiction and application of the different laws of potentially many jurisdictions in a massive network of constantly communicating computers spread among most of the countries of the world creates an instant Conflict of Laws nightmare. Such action would also induce similar actions by multiple jurisdictions that would lead to ever more complexity and conflict.
Because Internet traffic is dynamically routed in packets over multiple networks and paths, it's not feasible for any operators to be aware of the traffic content. In many cases, this inability to be aware of or to control content also applies to the services provided on the attached host computers. Imposing responsibilities on such providers for the passive passage of traffic would be utterly vicarious.
Lastly and perhaps most importantly, the Internet is simply a human communications medium, and all the existing civil and criminal law throughout the world still applies. Indeed, with the Internet increasingly interoperating with other media such as fax and even voice telephone services, it's becoming meaningless to distinguish among media. Civil and criminal actions have been brought in matters involving espionage, tort, libel, fraud, distribution of obscene materials, among others. Apart from the intractable problem of a lack of rules for resolving Conflict of Laws in this arena, it appears that existing law would suffice.
At present there are about 7 million computers indicated as reachable on the Internet. The growth has been consistently doubling every year for the past several years. Conservative projections based on the actual average growth over previous three years indicate about 120 million connected machines at the end of the decade. Although North America has the largest number of connected computers, the trends are quite amazingly similar in other regions of the world - even if the numbers are smaller. In general, the aggregate growth outside the USA is greater than inside the USA, thus assuring a continual globalization of the network.
It seems likely that these growth projections will materialize. The optimism derives from a number of developments now underway that all converge to further the Internet phenomenon.
A major computer system vendor has publicly noted that computers - as a communications technology - are growing and diffusing worldwide faster than any previous communications medium, including telephones, television, or VCRs. There are now about two hundred million computers and the number proceeds inexorably upward. Thus there are a constantly expanding number of machines that are potentially able to be Internet connected.
The ability to connect a machine to the Internet is affected by several factors: the physical components in the computer, the operating systems, and the ready availability of cost-effective access service. In the past year, good high performance network cards and modems have become so inexpensive that they're routinely shipped with increasingly large numbers of computers. Even more importantly, the Internet has become so popular that the necessary access software has been included with virtually all new computer operating systems, and large numbers of companies worldwide in almost every form and using almost every kind of access medium have become Internet Service Providers of network access.
The stage for entering the Internet provisioning business has been set by an array of telecom- munication and trade policies that have liberalized the use of telecom networks, driven the price of circuits down, and eliminated or diminished restrictions on providing enhanced or value added network services - the regulatory category into which the Internet falls. The bottom line - it's easy to enter the Internet access business, and to offer customers high performance at low cost - usually at flat rates based on the access bandwidth provided.
Meanwhile, a combination of innovative genius that resulted in such developments as the World Wide Web (WWW) and the Mosaic browser, as well as elegant user-friendly point-and-click tools inspired by the growing market, has enabled almost anyone over the age of five to use at least some of the Internet’s capabilities.
The development of new tools has not stopped. With almost each passing day, new techniques, new software, new capabilities become available - usually on the network itself. The Internet has become the ultimate computer development engine for producing new applications, while all of its non-"geek" users avail themselves of a constantly expanding array of new capabilities to better collaborate, educate, do research, share information, market, sell, and correspond. Perhaps underlying the Internet phenomenon is an intrinsic desire of people in the world to interact with others - near and far. For many families with college-age members, it has become a vital link between parents and children. For anyone with a global perspective, profession, business, or life- style, the Internet has become the most unifying personal medium of our age.
These services provide tens of millions of people with capabilities such as information sharing, global collaboration, distance education, software distribution, scientific research, product devel- opment, public services, marketing, sales, customer support, professional development, correspondence, and entertainment.
Only a minute fraction of this traffic is even potentially related to the transfer of objectionable materials, and recent assertions of significant transfer of "Cyberporn" on the Internet are totally misleading and contrary to the facts - evincing a profound ignorance about what even constitutes the Internet. See Response to Time magazine.
This is not to say that on some of the Internet's 7 million computers, there aren't some publicly accessible offensive materials, or that some messages don't convey such materials. As the Internet continues to scale to encompass an ever larger slice of the general public worldwide, it is inevitable that some materials will exist that some individuals or groups will find offensive. It's just that the numbers are proportionally very small and likely to remain that way. It's also worth emphasizing that a significant amount of such material is actually made available via dialup computer Bulletin Board Services (BBS), and not the Internet.
The most recent and largest scale of these activities was a special session last Wednesday at the 33rd Internet Engineering Task Force (IETF) meeting at Stockholm. The IETF is the international standards organization of the Internet. The special session was chaired by Internet pioneer Vint Cerf, and brought together several dozen experts from around the world to discuss their concerns and potential solutions. The occasion also provided opportunities for separate meetings and demonstrations of ongoing work. The session resulted in a proposal to create a new IETF Working Group directed toward developing necessary technical standards over the next six months.
In addition to the IETF action, the World Wide Web Consortium in Cambridge MA also announced its establishment of technical standards initiatives focused on WWW-based services. It was also apparent that several vendors have implemented initial products and that product development was underway in at least two research establishments.
The work is divided into four broad areas. The first - host access control - has actually produced products now on the market. These are effective, but have poor granularity (i.e., entire sites are either included or included), and are difficult to maintain because they require constant searching and updating by the vendors.
The second approach (also called 2nd party labeling) potentially allows individual files or pages of materials to be rated and filtered. It also relies upon the source to label the materials, and thus is easily maintainable. Non-labeled materials can be excluded.
The third approach allows 3rd parties to maintain labeling services. Thus one could potentially subscribe to a favorite commercial, educational, church or political rating service. However, this transfers a significant maintenance responsibility to that service. See the LANL testbed.
The fourth category is a set of approaches that rely on different existing effects of techniques to limit access.
The Committee should note that the second or third approaches invoke some ancillary public policy considerations. For source labeling to be really foolproof, it will require the use of encryption- based authentication technology now subject to export controls and usage restrictions in some countries. The third approach - 3rd party labeling - will require some certification of institutions to be foolproof, and a role for the Patent and Trademark Office might be considered to assure that only one legitimate rating organization with the same name exists.
Much of this work has an additional positive benefit of bringing about more efficient discovery of information, and possibly assisting in the enforcement of copyright claims. On the negative side, however, concerns exist about the abuse potential of the tools by those intent upon institut- ing political, religious or cultural-based controls on a population, as well as those who might purposely select objectionable materials.