------------------------------------------------------------------------ To: Information and Communications Services Policy Group From: Electronic Frontiers Australia Inc. Subject: Submission in response to "Consultation Paper on the Regulation of on-line Information Services" 7 July 1995 Date: 28 August 1995 ----------------------------------------------------------------------- Contents Introduction & Summary About EFA Answers to Secific Requests for Comment Regulation Technologies Recommendations Appendices Internet Parental Control Frequently Asked Questions Parental and User Control Technologies Available Today to Screen Unwanted Content -- Introduction & Summary -- This submission by Electronic Frontiers Australia (EFA) is in response to the "Consultation Paper on the Regulation of On-Line Information Services" of July 1995. It is based upon work done by a joint EFA/ACS (Australian Computer Society) Task Force. The submission starts with a description of EFA, followed by ... It makes several points, the principle ones being: o Electronic Frontiers Australia are pleased that the Government has issued the consultation paper electronically and has provided an email address for comment. There are several good points in the paper, however the main proposal of regulation will not prevent access to "contentious" material. o EFA think that it is a mistake to assume that the best and only solution to the problems is regulation. Technical solutions available and under development are not considered in the paper. EFA are opposed to the principle of regulation because no amount of regulation short of cutting Australia off from the rest of the world will work. Infact many of the reasons why regulation won't work are mentioned in the consultation paper. o The emerging Internet Service Providers in Australia are not information services who can control the content that their users access. In fact they have less control over the content than the users. Short of pulling the international plug no system of self regulation or government imposed regulation will control anything but a very small proportion of the content that users have access to. o The type of regulation proposed is beyond the territorial ( at least ) competence of the Australian governments and is therefore folly from the outset. On the down side regulation will impose costs and risks on service providers and damage Australia's commercial interests. o The addressing of concerns as to minors' access is better served by encouraging parental and teacher supervision, and in due course by the use of next-generation web filters. Analogies with purchased or broadcast media are irrelevant. o As the consultation paper notes what happens now will have a major impact on how these issues are dealt with as new broadband services (components of the Information Superhighway) emerge. It is important that we get it right now. Inappropriate regulation could cripple the potential of broadband services. o There is a parallel with another embryonic industry that well meaning censorship attempts by Australian Government effectively killed, the FILM Industry. Is it not possible that with the current Multimedia $M budgets the left had has lost sight of the right here? Censorship CAN and HAS ensured Australia not only missed the boat BUT exported its best and brightest to work in the field but under other flags. Then paid others for imports to see that which regulation prohibited from being made in Australia. The same CAN happen again. -- About EFA -- Electronic Frontiers Australia is an association of people with a common interest in computer based communication systems, online information services, electronic mail networks, and similar media. "Born" on Fidonet and the Internet last year, EFA arose out of a growing awareness by its founding members of the need for a community based organisation to become involved in the establishment of the national agenda for the delivery of data communication services to and within the Australian community-at-large. It is loosely modelled on the American Electronic Frontier Foundation (EFF) and received much encouragement and support from that organisation in its establishment phase. Electronic Frontiers Australia is a net-based organisation. EFA conducts the vast majority of its business via electronic data communications medium, utilising both Internet and Fidonet systems. EFA's objectives, loosely stated, are; - To ensure that people have the same basic freedoms 'within' computer based communication systems as without, - To educate the public at large about computer based communication systems and their use, - To support, encourage and advise on the development and use of computer based communication systems (and related innovations), and - To research and advise on the law as applied to computer based communication systems and related technologies. To these ends, EFA is corresponding with the media, representatives of the government, the law community, and other interested groups. EFA has created and/or uses mailing lists, USEnet newsgroups and Fidonet- based echomail conferences to foster discussion of these issues. -- Answers to Specific Requests for Comment made in the Consultation Paper -- The Consultation Paper requested comments on eight specific points. These points are treated in the following eight sections. "Whether the overall scheme is appropriate for regulating on-line information services" The proposed scheme attempts to maintain a consistent standard of censorship across all media. However any scheme that does not take the technological mechanisms and limitations of a particular medium into consideration will not be enforceable or effective. Legislation can not be appropriate if it is not both enforceable and effective. In order to evaluate the enforceability and effectiveness of the proposal, this response will test it against the five intentions specified in the Consultation Paper: * aligning censorship regimes for new services with the regimes that have been adopted for other media. There are major technical obstacles to aligning Internet censorship with the regimes adopted for other media. Some of these are: . Material cannot be classified before it is distributed on the Internet. The Internet is an interactive medium - it consists of a great number of real-time conversational forums. If material had to be classified before being distributed, Australia would not be able to participate in these real-time interactions. If material were to be classified after it had been distributed, then individuals would not know whether their material might breach the censorship standards until it had already been disseminated. . Most of the information accessible on the Australian Internet is served from foreign sites which cannot be controlled by Australian laws or regulations. . No regulatory body can keep up with the volume of Internet communications. This volume presently exceeds 30 terabytes per day on the global Internet; an ISP only propagates a tiny fraction, perhaps one ten thousandth of this material, but this still constitutes several gigabytes (a gigabyte is about a thousand bibles) of material per day. . Information is propagated on the Internet with extreme rapidity. It is not unusual for a news posting to be copied all over the world and read by many hundreds of thousands of people within hours of its origination. No regulatory body could be expected to act with sufficient speed to prevent the broad dissemination of unclassified material originating from Australian Internet sites. . No centralised body can have access to or receive copies of all the material propagated over the Internet. The Internet is a multicast system, not a broadcast system, so different sites propagate different messages, and different sites receive different messages. The Internet is entirely decentralised, with no center or root through which all traffic must flow. Given these fundamental technical properties, standard regulatory regimes can have no effect on material accessible via the Internet in Australia. However, as we describe below, an approach that recognise the technological limitations of the medium, teemed with technological solutions, can be quite effective and are highly desirable. Detailed recommendations for such an approach are provided below. * protecting freedom of expression, especially with regard to private communication between adults; To protect freedom of expression, adults must be indemnified against legal penalties that arise from expressions misattributed to them. However, at present, it is easy for any Internet participant to forge or misattribute expressions of others, both in email and on the worldwide netnews or "USENET" multicast system. Such forgeries are often constructed so that they are technologically impossible to detect. Any Internet user from anywhere in the world may perpetrate such forgeries, and any user can learn how to do so simply by reading the publicly available Internet documentation. The Consultation Paper does not consider the implications of such forgeries; the legislation that it proposes would be susceptible to terrible abuses by criminals and pranksters perpetrating this kind of fraud. * limiting children's exposure to harmful or unsuitable material; No Australian regulatory solution can be applied to the foreign Internet sites which serve most of the material available on the Australian Internet. On the Internet, foreign material is no more difficult to access than local material; an Internet user is usually not even aware of the geographical location of the material that has been accessed. The proposed scheme will not prevent the exposure of Australian children to harmful and unsuitable material that derives from foreign Internet sites. Happily there are technological solutions, including those proposed by General Products, Netscape and Microsoft, described below and in the appendices to this response, that provide the technological support necessary for a solution to this problem. * supporting the development of new services that enhance Australia's competitiveness; The proposed regiem makes it possible for local Internet Service Providers (ISPs) and IT professionals to become liable for information that is stored on their machines. It provides them with a range of defences against such liability, but it fails to account for the financial impact of equipment seizures, violations of corporate security and expensive legal defences that will also result from the proposed legislation. Such impacts may greatly affect the competitiveness of Australia's multimedia and IT industries, neither of which can survive without the ability to communicate, advertise and provide services on the global Internet. Any Australian Internet participant could be harrassed and blackmailed by criminals and pranksters under the proposed legislation. No business can afford such risks. As Australia's IT and multimedia industries cannot compete without Internet access, they might even be driven out of Australia under the proposed legislation. * ensuring that emerging service industries are not burdened with unnecessary costs; The emerging service industries are not merely concerned with cost - if there is a possibility that their customers may, through no fault of their own, be forced to defend themselves against criminal liability for material that is transmitted to and stored on any of their machines by foreign users, or which is attributed to their customers as the result of forgeries and other frauds, then it will become extremely difficult for these service industries to conduct business in Australia. In summary then, the proposed policy will be neither effective nor enforceable. Its effect on Australian service industries will be devastating. It is open to terrible and widespread abuse. It is, therefore, not appropriate for regulation of the Internet. "if it is not appropriate, any other available means for regulation of BBS" The Consultation Paper confuses BBSes, which comprise a few thousand teenagers and hobbyists, with the Australian Internet community, which comprises almost a million Australian citizens. The Australian Internet community includes most Australian scientists, engineers, artists and academics, as well as most large and medium sized Australian businesses. Within five years the Australian Internet will be more commonplace than fax machines. Its concerns are central to Australian commercial and cultural pursuits. Although BBSes and the Internet have in the past used different technologies, there is a growing convergence. Already some BBSes provide access to the Internet and this trend is likely to continue. Since there is no clear dividing line between BBSes and Internet access providers, this response concentrates on the available means for regulation of the Internet. As we have noted above, the legislative solutions proposed for moderation of Internet material are not appropriate. It is very unfortunate that the Consultation Paper does not cover technological solutions to the issues it raises, for such technological solutions are certainly not infeasible, nor is there any reason to think that these solutions won't continue to be feasible as the technology develops. Some of the technological solutions that are being implemented to regulate the Internet are described below and in the appendices to this response. These solutions can not be made compulsory across the Australian Internet community, and so are not in line with the censorship regime that Australia requires of other media. However these technological solutions present an effective solution to many of the concerns laid out in the Consultation Paper. They are the only solutions that appear to be practicable without susceptibility to the difficulties described above. "whether the definition of "on-line information service" is too narrow or too broad" It is not correct to classify BBSes and the Internet under the same category, due to the size and character of their participant populations. But there are also qualitative differences between these technologies that must be taken into account when regulation is considered. The qualitative difference between BBSes and Internet sites may be thought of as the difference between a solid and a liquid. Like a solid, the information on a BBS does not flow. BBS files are uploaded and downloaded only by the BBS's users and administrators, all of whom may be registered with and monitored by the BBS. It is a relatively simple matter to regulate these users and attribute responsibility for the information that appears on the BBS. Like a liquid, the information on the Internet flows through many hundreds of thousands of machines. It does not remain in local files, but is piped to and from access programs that are distributed on computers all over the planet. It is never possible to be certain of the location of a particular piece of information, nor of the identity of its originator. It is a commonplace and simple matter for all users to access perfectly secure and untraceable anonymity and encryption services as described in Appendix B. It is almost never possible to be certain of responsibility for the information that is available via an Internet node. Thus individual BBSes may be shut down, but the Internet cannot be shut down; it simply flows around any sites are unavailable. BBS users may be deregistered, but Internet users may maintain dozens of accounts in dozens of countries; these accounts are often available for free and without requiring any authentication. BBS files may be deleted, but information propagated on the Internet is copied to many different countries within hours of its appearance, and all of these copies are directly available from any Australian Internet node. So the definition of "online information service" is neither too narrow, nor too broad - it is simply nugatory. BBSes and the Internet cannot be regulated under the same policy, and it is a fundamental error to attempt to treat them as two examples of the same concept. The definition of "online information service" as given in the Consultation paper also covers Cole's cash register. In attempting to have a definition which encompasses both the amateur BBS and commercial internet providers, the context has been lost. By all means consider file transfers from BBSs or overseas web page access from Australian internet sites, but the two must be treated differently. There is a wide range of online services, with different levels of control by operators, and different potential abuses. "Online information service" is a poor choice of name. Calling electronic networks "on-line information services" is like calling a telephone network an "audio information service" because 0055 numbers can be accessed using the network. To take the analogy one step further it would be inappropriate to call a company providing local exchanges and exchange connections an "audio information service". Similarly it is inappropriate to call an Internet Service Provider an "online information service". "the role of the education strategy in informing parents, teachers and children about on-line information services and the ways of addressing possible problems" The Internet is larger, busier and more populous than any city on Earth. Like a city, the great majority of its resources are provided for educational, commercial and recreational purposes. But also, like a city, it has a few seedy dives, a small but readily available collection of strip joints and bordellos. Unfortunately, these areas hold a fascination for our broadcast media, who have presented them as representative of the broader Internet community. This is a gross distortion. 99% of such dives exist outside of the Australian jurisdiction, so there is nothing that can be done to close them down; as with city life, we can only keep a close watch our children, keep them close to places that we know well, and educate them about the wider world. If we mistakenly introduce legislation that is destructive to our own portion of the vast global Internet, we do our children a far greater disservice even than letting them encounter the "online King's Cross". We doom them to uncompetitiveness and ignorance of the fantastic opportunities that the Internet provides. The global Internet is simply beyond Australia's ability to regulate. Therefore it is vitally important that parents and educators monitor and restrict the activities of their children themselves. Tools with which they may do this are described below and in the appendices to this response. "non-criminal sanctions that could be applied for non-compliance with a code of practice" The information distributed by Australian Internet sites is not originated by ISPs, but by Internet users all over the world. ISPs have no control over the actions of these users, and so no code of practice can be devised for ISPs that will affect the information provided by Australian Internet sites. An obvious possibility is that ISP customers (Australian Internet users) themselves could be asked to adhere to a code of practice. But because most of the Internet is owned and operated outside of the Australian jurisdiction, such measures would have virtually no effect on the information available at Australian Internet sites. "a complaints mechanism and possible establishment of an independent complaints handling body" Judging by the debates on the subject of regulation that have occurred on the Internet over the last year, it appears that there would be great popular opposition to such a body, which is generally thought of by Internet participants as the sort of "Big Brother" popularised by George Orwell's "1984". Why is this? Why should Internet participants be more opposed to such a body than, say, television viewers? The difference between Internet participation and television viewing is like the difference between writing and reading. Censorship of television does not infringe upon a viewer's freedom of expression, but censorship of the Internet does. Opposition by the Australian Internet community could easily sabotage a complaints body by generating large numbers of spurious complaints concerning innocuous material, as well as by clogging and disconnecting those parts of the Australian Internet that give connectivity to that body. And again, as most of the Internet is owned and operated outside of the Australian jurisdiction, it seems likely that most genuine complaints handed to such a body could never be resolved. "issues arising from the practical implementation of the proposed offence provisions" New users are arriving on the net at an astonishing rate. The Internet has doubled in size every nine months for the last ten years; it presently includes over fifty million participants, including most of the scientists, engineers, artists and academics on the planet. Its rate of growth will not stabilise until it includes most of the computers in the world. The best estimates of MIT's Media Lab suggest that this will occur in around six years time, at which point the Internet will include better than eight hundred million (800,000,000) participants. This will makes the Internet the world's largest marketplace. The commercial possibilities, especially to a country as geographically isolated as Australia, cannot be over-estimated. Any laws that are enacted by the Australian government must be formulated so as not to hinder Australian commercial and cultural interaction with this tremendously wealthy and diverse global community. The consequences of such poorly framed legislation will be economically devastating to this country. The great bulk of the information transmitted through and available from the Internet has not been classified. Nor will it be feasible to classify this information, due to its massive volume. The quantity of information that passes through an ISP each day is far beyond the capacity of its users to scan by eye, and there exist no automatic programs capable of scanning it - it is too complex for even the smartest AI to interpret. The transient nature of this information is also unique to the Internet. Every day over a million web pages appear, disappear, or are subject to change. These changes are not propagated automatically, so there is no way to screen them. In addition, a page may appear differently to different users, so that when one person clicks on it it serves up contenious material, but when a censor looks at it it serves up images of butterflies. "any other matters that may be relevant" The consultation paper is ten years behind the times. By the time the proposals could be implemented, technical solutions will emerge to both circumvent any legislation and provide parents with more control over the material their child can access. The desired result of protecting children can be achieved by empowering parents and teachers. Such an approach will be effective, regulation will not. Such an approach will not have adverse consequences, regulation will. If the government wishes to reduce abuse of online services, it needs merely to encourage more people to use them. Community standards are more likely to apply in a medium to which the whole of the community has informed access. -- Regulation Technologies -- Voluntary Techniques Simple techniques that have existed for some time include kill-files, mail-filters and gag-lists. A kill-file filters netnews according to particular keywords, allowing newsreaders to hide articles on particular topics, or written by particular authors; mail filters allow much the same thing to be done for email, as well as other simple transformations such as deleting swearwords and so forth; gag-lists are again the same idea applied to real-time interactions on MUDs and IRC. These techniques are purely voluntary - a user can apply or remove them at will. They are generally effective on written media, but do not work with world-wide web pages that contain sounds, pictures and movies. They also do not work until the user has already been exposed to some offensive material. Filter Technologies The two appendices to this response give details on some filter technologies which are currently available or under development. The remainder of this part of the response is devoted to specific examples of such filter technologies. SurfWatch and NetNanny Two corporations, Trove Investments and SurfWatch Software, have produced products that enforce censorship on children. These products are "NetNanny" and "SurfWatch". These products are described below; it should be recognised that they are not effective means for forcing censorship on adults, who would readily devise and acquire software and techniques to circumvent them. NetNanny scans all incoming and outgoing communications for keywords selected by an adult - a parent or educator - and filters-out or disconnects from Internet resources that include those keywords. This scheme relies on the ability of the parent or educator to foresee all possible objectionable keywords; that's pretty hard to do. NetNanny is unable to deal with simple transformations of text, which are easily applied by users, and it cannot detect or monitor objectionable sounds, pictures or movies. For these reasons it is not always effective. SurfWatch is a little cleverer. As well as using a set of keyword filters, SurfWatch employ a team of human censors who monitor the web and declare particular resources to be objectionable; the SurfWatch software downloads this list and uses it to block access to those resources. This confers two advantages; it takes a lot of work away from the parent, and it can deal with pictures, movies and sounds. Although it doesn't provide much protection against objectionable email, netnews or realtime interactions, it is mostly effective. IHPEG Recently Netscape and Microsoft have announced that they will cooperate in a consortium called IHPEG (Information Highway Parental Empowerment Group) to produce a completely effective solution for providing children with a censored view of the Internet. IHPEG software, to be released later this year, will allow parents and educators to cooperate globally to make the Internet safe for use by children. Netscape's scheme allows parents and educators to tag particular resources as being "kid-safe". These kid-safe tags will include digital authentication of the variety described in Appendix B, so that if the resource to which they are attached changes, they will be invalidated. Children will then employ the Netscape Internet access software to interact with the net in perfect safety - they will not be able to read anything that has not been deemed suitable by an officially registered parent or educator. -- Recommendations -- * It should be acknowledged that regulation will not be effective. * No new legislation should be introduced. * Technical solutions should be promoted. * A parental and teacher empowerment program should be set up to provide parents and teachers with access to filtering technology. -- Appendices -- Reports by two US organisations, Voters Telecommunications Watch and The Center for Democracy and Technology, are appended to this response. Some of the details given in these reports are specific to the United States, particularly legisative details and prices, which are in US dollars. However most of the information given is of a general nature and is just as applicable in Australia as it is in the US. Internet Parental Control Frequently Asked Questions [posted separately] Parental and User Control Technologies Available Today to Screen Unwanted Content [posted separately] -- Dr Michael Baker, EFA Board Chairman PO Box 5, Flaxley, SA 5153, Australia Internet:mbaker@apanix.apana.org.au Ph:08 388 8439 Fax:08 398 2250 [08=+618] Fidonet:Michael Baker, 3:800/838 For EFA info email efa-info@efa.org.au Compuserve: 100026,1321