AustLII Home | Databases | WorldLII | Search | Feedback

Privacy Law and Policy Reporter

Privacy Law and Policy Reporter (PLPR)
You are here:  AustLII >> Databases >> Privacy Law and Policy Reporter >> 1996 >> [1996] PrivLawPRpr 60

Database Search | Name Search | Recent Articles | Noteup | LawCite | Help

Kirby, Michael --- "OECD Cryptography Guidelines in context" [1996] PrivLawPRpr 60; (1996) 3(7) Privacy Law & Policy Reporter 121

OECD Cryptography Guidelines in context

Justice Michael Kirby

As the OECD draws closer to finalising its Cryptography Policy Guidelines (see extracts from a draft in this issue), developed by an Ad-hoc Group of Experts chaired by Australia, Justice Michael Kirby sounds a timely warning of the dangers and difficulties in developing international policies in such emotive areas as cryptography and HIV/AIDS. This paper was presented by Justice Kirby at the International Symposium on the Public Voice and the Development of International Cryptography Policy, Centre De Conference International, Paris, France, 25 September 1996. The Conference web site ( also includes a report on the proceedings of the Conference and articles from Le Monde. (General Editor)

A nostalgic look back

We are in Paris once again. If I close my eyes nearly twenty years roll away. I am chairing the OECD Expert Group on Transborder Data Barriers and the Protection of Privacy.

It is important to view the OECD work on cryptography in the context of its lineage. It is a project which follows directly upon the recommendation of the Council of the OECD of 23 September 19801 commending to member governments the Guidelines developed by the first Expert Group. It also follows the recommendation of the Council of the OECD concerning the Guidelines for the Security of Information Systems, adopted on 26 November 1992.2 But more fundamentally, the new Guidelines will, in due course, be proposed to the Council of the OECD, a body established by its Convention3 with the economic focus that its title suggests but agreed between nations which have certain precious things in common. These are a commitment to the rule of law; an acceptance of the democratic system of government; and a respect for fundamental human rights.

In 1960 when the OECD was established, in the midst of the Cold War, the reference to these specific features of OECD countries did not need to be over-stated. They were just assumed. They were the things that made the OECD a unique and special place: quite different from other international bodies. This was not only just a club of the developed and other countries of the world, exchanging information vital for their continued economic success. It was also an organisation whose members accepted certain governmental norms and principles. It is important to remember these features of the OECD for they provide the milieu within which the cryptography policy guidelines must be fashioned for presentation to the Council of the OECD, still representing the governments that share these common features.

It is worth recalling to mind, just for a moment, the remarkable and talented groups of people who made up the two earlier Expert Groups. They included some of the most intelligent people I have ever worked with. Many of them have gone on to distinguished national and international service. Mr Louis Joinet became an adviser to the President of the French Republic and is one of the Special Rapporteurs for Human Rights of the United Nations. Mr Hans Correll of Sweden went on to become the General Legal Counsel to the United Nations. The same high talent was seen in the second group. Both groups were well served by the OECD Secretariat, led by Mr Hans Peter Gassmans, ably supported by Professor Pieter Seipel and Mrs Deborah Hurley. It is not an exercise in nostalgia to say this. It is a reminder of the fact that the high success of the earlier Guidelines derived from work of people of considerable talent, a strong sense of independence and a good understanding of the context of international human rights law within which they were drafting their Guidelines.

I do not pretend that the battles over human rights issues do not sometimes have an economic dimension. Thus, in the first Expert Group, the commitment of the US to free flow of information was said to draw strength not only from that country's political and legal culture but also from its then dominance of information technology. The European countries commitment to privacy protection drew on the experience of those countries in the misuse of information systems during the war. But it was also said to take strength from a desire in Europe to promote a local information industry. However that may be, the experts were able to reach the necessary compromises. The Guidelines were agreed. They were recommended to, and adopted by, the Council. And all this was done in the usual way of the OECD. Patient consensus building rather than impatient demand for action is the way that the OECD operates. A measure of the success of the Guidelines, 16 years later, is the extent to which they have been adopted, virtually unaltered in the domestic law of many members countries of the OECD, including my own.4 The Guidelines for the Security of Information Systems have also proved influential in many countries, providing the impetus to the consideration of local laws and policies.

It is self-evident that this history of the development of Guidelines, in the context of international information policy, is a most precious resources of the OECD. It is respected and valued in the member states and beyond. It will be important that nothing be done in the preparation of the Cryptography Policy Guidelines which diminishes the high reputation which the OECD Expert Groups enjoy in this field both for their product and for their methodology.

This needs to be said because, in the field of cryptography, as in the earlier areas of privacy and security, a failure of the OECD's efforts, or attempts of individual member states to `go it alone' are likely to provide ineffective, inefficient, such as will impair international trade and the development of the Global Information Infrastructure and the Global Information Society. These goals are important not only for future economic well-being of OECD member states but of the whole world. Moreover, they are part of the definition of the future of our world in which human beings everywhere will flourish and enjoy the stable, ordered, democratic, and rights-respecting environment which the current members of the OECD generally enjoy. It is inevitable that the Council of the OECD will want to hand that legacy on to future generation, enhanced and not diminished by advances in information technology. All of this may be accepted. The question presented by the development of information technology relevant to cryptography is whether these have changes the `playing field' such that new powers must be afforded to the state to monitor and intrude into information systems, including those which have been encrypted to protect integrity, confidentiality, privacy, and other qualities of information.

I do not pretend to a detailed expertise in the issues which the cryptography guidelines will be present for debate. But in the 20 years since I first laboured over OECD Guidelines, I have had much experience in several international and national bodies relevant to human rights. Back in 1978, the OECD `house' felt a little uncomfortable in tackling the issue of privacy, so manifestly an issue of human rights. It was not the usual fare of economists, technologists, and government officials who walked the corridors of the OECD. But since then, stimulated in part by the achievements of the earlier experts groups and by a realisation of the very things which bind the OECD together, the issue of human rights have been assuming a higher agenda in the OECD, and rightly so. Not only do human rights have strong economic implications. They are the cement which binds together the member States of the OECD and provides the rationale of the Organisation: by economic cooperation and development to strengthen the social and individual environment within which the citizens of the OECD countries live and work. After all, economics has no point in itself, save as it serves and enriches the lives of individuals.

Lesson from a different world

Now let me take you into a different world. It is the world that has brought me to Europe on this visit. Recently, I have moved out of the field of information policy to other pressing concerns of the international community. One of them is the Human Genome Project -- the greatest international scientific co-operative project in history.5 But another is HIV/AIDS. Each of these projects represents a development that is of enormous economic potential, including for the member states of the OECD. Each poses important challenges to human rights. Each must develop within the context of international human rights law. Each presents highly controversial problems which are extremely difficult for democratic societies to solve without the help of informed experts and international co-operation.

The immediate reason for my visit to Europe is a meeting at the Palais des Nations in Geneva for a consultation on human rights issues of HIV/AIDS summoned by the High Commissioner of Human Rights and UNAIDS -- the joint program on AIDS established by the Secretary-General of the United Nations. I express my gratitude to them publicly for releasing me from a morning session of their consultation so that I could come to this symposium in Paris. They agreed that the issue you are addressing is also of great importance to the international community and to human rights. That is why they let me come.

Ten Commandments

Permit me to draw upon my work of the last decade in policy development on HIV/AIDS, including the WHO Global Commission on AIDS, to extract a number of lessons which, I believe, provide a framework within which cryptography policy guidelines should be developed by the OECD Expert Group. Let me suggest that there are ten commandments.

1. Accurate technical data

The first rule I learned in the provision of advice on global strategies to meet the challenges of HIV/AIDS was to rest all policies and law upon sound scientific data. Not hunch or guesswork, prejudice or good theories. But a sound understanding of the virus, its modes of transmission and its real challenge to the international community. It is the same with cryptography. The technology is moving rapidly: from Clipper to Capstone. From Fisher Watchdog, Entrust, Shoplock, Secure through Key Escrow, Weak Encryption, Link Encryption, and Strong Encryption. The starting point for devising the OECD Guidelines must be an absolutely accurate and up to date understanding of what technology can deliver where the technology is going and what implications these developments have, relevantly, for national security and law enforcement (on the one hand) and privacy protection (on the other).

2. Avoid prejudice

In the struggle against HIV/AIDS the word is full of prejudice, alarm, hysteria, and exaggeration. At the beginning of the pandemic, many states adopted strong criminal sanctions against people infected with the virus. A few locked them up. The calls for extreme measures were the siren songs of local politicians, responding to worried citizens demanding immediate action. Many such responses were completely ineffective because of the shifting target of the pandemic and the extreme difficulty of influencing personal behaviour in the myriad of manifestations. I suspect that encryption, even if it does not present an inevitable crypto-anarchy6 presents similar problems to those drawing up laws and policies (and guidelines) on cryptography. Out there in the cities and towns are millions of minds devising methods of cryptography, just as there are millions of people engaging in risky behaviour. There should be a clear sighted understanding of the limitation of law in influencing such behaviour. Exaggerated notions of the law's role can only lead to over-reach and failure of the law's response.

3. Cultural differences

Another lesson in the struggles against HIV/AIDS is that international norms are very difficult to attain because of differing cultures, histories, and legal postulates. I suspect that this is the same with cryptography except that to this mix there must also be added trade and economic advantages. In the struggle against HIV/AIDS it has often been necessary to opt for levels of agreement that are less than perfect simple because of great divergence in local outlook and interests. This may be true in this area as well.

4. Sound methodology

Another lesson from the field of HIV/AIDS is the need to involve, at every stage, the consumer groups who will be affected by laws and policies. This is why a high level of transparency has been adopted by international agencies such as UNAIDS and by the governmental programs of those countries, like my own, which have actually made an impact on containing the epidemic. I think that this is important in the field of cryptography as well. I applaud the involvement by the OECD of trade, industry, telecom, law enforcement, national security, private sector and date protection agencies in the work of the Expert Group. But it will also be vital, if the program is to be a success in practice, closely to involve these bodies that speak for values which may sometimes be in competition. One of the real successes of the international and national work against HIV/AIDS has been the involvement of the non-governmental organisations who speak up for the infected, their families and careers. It is equally vital in the work of the Expert Group on Cryptography that its product should be exposed before completion to the critical attention of civil society organisations which speak, for example, for privacy, for individual rights and for the containment of the power of surveillance by organs of the state. If it is not done by the Expert Group one thing is sure. The Guidelines, when returned to the member states with the recommendation of the Council and the OECD, will fall into the net of community debate about privacy and other concerns. If real action on the OECD Guidelines is sought, it is vital that strong voices for the relevant human rights should be expressed at the OECD table. This was done in the earlier Experts Groups, often by the experts themselves and sometimes by invited observers. There would be a danger if the issues of cryptography, for example, were turned over to the viewpoint of law enforcement and national security representatives (however important) to the effective exclusion of advocates for privacy and democratic rights.

5. Language and symbols

One thing has been learned in the struggle against HIV/AIDS and that is the importance of language and of symbols. Especially against the background of past epidemics and the highly ineffective but oppressive responses of nations and of the international community, there is often alarm at the dangers to individual rights which can actually be counterproductive in tackling the problem. So too in cryptography. It is important there should be no under-estimation of the significance of the symbols which will be sent out by the Guidelines to the various audiences which will be scrutinising them in the free societies of the OECD. Thus, to delete references to privacy, or to regard them as adequately covered by the historical allusion to the previous Guidelines, would not be missed by those who champion privacy concerns. In the business of international guidelines, language and symbols assume a rare importance for they are the signals to action in the government to which the guidelines will be addressed.

6. Alarms in proportion

It is important that the Guidelines should be based not only on sound scientific understanding of the nature of cryptography and what is and what will be available. But also on a clear understanding of the social problems which are said to give rise to the need in government to break the encryption and invade it. Encryption is not the only protection for privacy; but it is one protection. The demands for the right to governmental intrusion into private messages and commercial secrets are usually expressed in very general language:

These cases involve child pornography, customs violations, drugs, espionage, embezzlement, murder, obstruction of justice, tax protesters and terrorism. At the International Cryptography Institute held in Washington in September 1995, FBI Director Louis F These suggestions paint opponents to the enhancement of government power into a corner as those favouring child pornography (at the top of the list), terrorism and the assassination of the Pope. Who could favour such things? But the lesson of the response in democratic societies to terrorism has been the importance of resisting extreme authoritarian measures in the attempt to combat it. When this happens, the terrorists actually win. The lesson from HIV/AIDS is that alarmist talk must be viewed with real suspicion and subjected to cold-eyed scrutiny to measure the real scope of the problem that is said to require extreme measures.

7. Tripartite test

Another lesson from HIV/AIDS is that the policies and strategies must be developed in the context of international law. This requires that they should respect fundamental human rights, such as the right not to be subjected to arbitrary or unlawful interference with privacy, family, home, or correspondence.8 With the advent of the HIV/AIDS epidemic public health officials demanded complete exemption from human rights restraints. However, it is now generally understood that those restraints must be kept in place. Any derogation from fundamental rights must be subjected to three limitations:

I suggest that the same limitations apply to each every demand by national and law enforcement agencies for a power to monitor information, including the subject of strong encryption. No individual or agency is above the law in OECD countries. Nor should they be.

8. Ongoing action

Another law is that the target of policy and law constantly shifts. The virus mutates. New areas of the globe with different problems are attached. So it will be with cryptography. The technology, that national interests and the skills and capacities of individuals will change rapidly. Already in significant respects the 1980 Guidelines on Privacy have been overtaken by new technological capacities of information systems. There should a healthy modesty in the views of the Expert Group as to what they can achieve, they need to recognise that the task in which they are engages is an ongoing one, although it presents immediate and urgent dilemmas.

9. Government role

There is an urgent need to capture the attention of government at the highest level. This is so in HIV and it is so in cryptography. Alas, in the former governments run away from responsibility. The problem in cryptography may be the opposite. The voices of national security and law enforcement agencies will generally be close to the ear of government. It is important that there be voices of equal strength to speak for human rights, the rule of law and protecting the privacy of citizens from the technologically enhanced capacity of the State to monitor their communications. In Australia, citizens with a concept of the rule of law have been shocked by the recent revelations of corruption and manipulation involving the nation's largest police force (the NSW Police Service). Last week a further inquiry began into the Federal Police because of allegations of corruption. I am sure that my country is not alone in this danger. Where trade secrets, governmental data and vulnerable systems are at stake it is imperative that those who claim the key to the kingdom of encryption should themselves be subject to constant and fearless scrutiny against the misuse of such large power.

10. Action for effectiveness

In the battle against HIV/AIDS there is a role for governments and their agencies to take action. But in a democratic civil society most of the effective action will take place amongst individual and community groups. The task is to define the limited role of government action and then to keep its agencies within that role. They should be subjected to stern legal discipline and constant scrutiny to ensure that the agencies remain our servants, accountable to the elected representatives of the people. It is a healthy democratic principle for citizens of democratic countries to remain sceptical of alarmist protectors. The history of this century has been one of the misuse of power. Technology now enhances the power of intrusion. In the name of human rights, including privacy, it is important that the power of intrusion, including official intrusion, be kept in strict check. I suggest that such checks can be fashioned by reference to my commandments, and doubtless others.


I offer my respects to those who have the privilege to participate in the OECD Expert Group which follows those I was honoured to chair. I know enough of the problem of cryptography to realise that is not easy of solution. The issue is not, as I think, one of `balancing' the legitimate aims of law enforcement and national security (on the one hand) and protection of privacy and human rights (on the other). Each of these objectives has a legitimate claim on governmental policy and action. however, the claims of national security and law enforcement agencies must be attained within a context of constitutionalism, the rule of law and respect for, and effective protection of human rights. This was recognised by the earlier Expert Groups. It is, I believe, a reason for the success of the guidelines which they produced. Those Guidelines were themselves the product of free citizens, working together in the harmonious community of the OECD. It is imperative today to retain the momentum and the approach which the earlier Expert Groups embraced. This is not only important for the reputation of the members of the present group or of the OECD. It is important for the good government and happiness of the citizens of the member countries.

Sometimes it is useful to see one's problems through the prism of a different experience. That is why I have been brought from another world (well actually Geneva) to offer these few remarks. For the affection for the OECD, forgive an old participant. For the recognition of the limited but legitimate role for national security and law enforcement, I plead an understanding of the exceptional case, under legal warrant, where this may be justified and even imperative. But for the demand for effective respect of human rights, and especially individual privacy, I make no apology. Such rights represent the ultimate common denominator of the OECD. They should find reflection in the new Guidelines as they did in those that went before. Nothing less will do.

The Hon Justice Michael Kirby AC CMG, Justice of the High Court of Australia, is President of the International Commission of Jurists. One-time Chairman of the OECD Expert Group on Transborder Data Flows and the Protection of Privacy, and of the OECD Expe

1. c(80) 58/FINAL.

2. C(92) 188/FINAL.

3. Convention on the Organisation for Economic Co-operation and Development of 4 December 1960.

4. Privacy Act 1988 (Cth).

5. The author is now a member of the Ethics Committee of the Human Genome Organisation and of the International Bioethics Committee of UNESCO.

6. D E Denning, `The Future of Cryptography' <3 PLPR 33>. See also G Greenleaf, `OECD Searches for Crypto-Consensus' <3 PLPR 21>; R Clarke, `Cryptography Issues in Plain Text' (1996) 3 PLPR at 24; P. Ford, `Information Security, Censorship and Privacy', June 1996; S. Orlowski, `The International Debate on Encryption and Public Key Authentication -- Its Impact on Privacy', unpublished paper.

7. Denning, above n 6, at 34.

8. International Covenant on Civil and Political Rights, art 17.

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback