First of all, we would like to thank you for the attention you pay to our Magazine granting our request for an interview.
The paper you presented during the recent seminar at the Castle of Priverno (Italy) will be, of course, the guideline of our talk. It was the first international conference on intelligence issues with the participation of field experts, journalists, politicians and also of former and current members of foreign Intelligence Services (even from the former "Eastern Bloc" countries).
Your presentation rightly focused on the importance the human element (a key-element in your view) still maintains. Even though, as some observers noted, the unquestionable evolution the intelligence world has undergone and is still undergoing, appears to be geared more towards the technical rather than the human aspects. We believe that your placing the human element, with its weaknesses and intuitions, in the centre place and your passionate defence of the importance (today greater than during the Cold War) of agent infiltration in order to encourage defections from the opposite field, are extremely interesting and significant. The twenty-first century intelligence will not be subject to technology, it will take advantage of it without being overwhelmed by it . We have well understood your view, so different from that of those who consider systems such as Echelon the only option for the future. In this context, how is it possible to improve co-operation between Services and between them and the other national administration bodies, in order to avoid missing good opportunities because of comprehension mistakes?

I think that the first thing I would like to say is that I much enjoy looking at your Magazine. "Per Aspera ad Veritatem" is a valuable instrument of intelligence collaboration in itself.
Let us start from the conference. I think the conference illustrates one extremely important point: all analyses, all researches, and it really does no matter whether at an academic level or within an intelligence community, require discussion. As soon as any academic sees the research of somebody who has been working in isolation and who has not discussed with colleagues and not been to conferences, preferably international conferences, will see that the work suffers from a degree of isolation. In other words researchers in any field could not possibly refine their views without having the opportunity to discuss them with colleagues in the same area and that is I think why Priverno was so important.
Your question addresses the problem of improving co-operation between Services and the other national administration bodies in order to avoid missing good opportunities because of comprehension mistakes. Well, I think that one of the first things that I would say about that is that, as many other problems which affect intelligence communities in all countries, these are not simply problems for the intelligence, these are problems for policy makers and there is no prospect of really satisfactory solutions being devised for these problems unless policy makers take a direct interest in them. What I think has been very striking is that most of the literature on intelligence services, most of the discussion about intelligence reform, has been based on the false assumption that the only ones which need to be reformed are the intelligence services. Now, the most comprehensive study that we have on the current and future role of any of the world's major intelligence communities was produced in the United States five years ago (you will find it quoted in my paper of Priverno) that is the report of the Aspin-Brown Commission on "the role and capabilities of the US intelligence community". The first of the measures it proposed to improve what it called "the performance of US intelligence" was directed at policy-makers first, rather than at intelligence agencies. It said: "intelligence needs better direction from the policy level regarding both the roles they perform and what they collect and analyse. Policy makers need to appreciate to a greater extent what intelligence can offer them and be more involved in how intelligence capabilities are used" (that is the end of quotation and you will find it in my paper).
Now I think there has been a widespread assumption that the so called intelligence failures, that is the failure to produce or to use intelligence well, have been the fault of intelligence agencies. There have of course historically been many occasions when intelligence agencies have been at fault, but looking back on the history of the 20th century, I think that it is now clear that intelligence failures were actually more a result of the inability of policy makers to make good use of intelligence, than the collection of intelligence in the first place.
Your question also addresses co-operation between intelligence services. Now it seems to me that this is an obvious necessity. Both Italy and UK see their security within an alliance framework, we are both members of NATO and it is simply impossible to imagine that either Italy or the UK could look after their national security other than as partners of an international alliance. I think, therefore, that if we have to have the best intelligence, that involves collaboration with other partners. At the basis, I think, there is however a problem which is a structural problem and quite an embarrassing one, but it is a real problem and just because it is embarrassing that does not mean we should ignore it or not discuss it. That is: in an intelligence multilateral alliance, secrecy so far as sources and methods are concerned, is as strong as its weakest link. This means, I think, that the future of intelligence collaboration is much more in bilateral relationships, such as that between Britain and Italy, than in multilateral ones, and it is an embarrassing and nonetheless significant fact that not all members of the European Union keep secrets as successfully as others. I think the same applies to relations between intelligence agencies, as to the relations between human beings. In other words, confidence between human beings, the confidence that you require to take somebody else into your confidence, is not something which is produced at an instant, at a single meeting, it is something which is built up over a period of time, that is why we now talk of "confidence building measures". So, I see the future of European intelligence very much in international collaboration, but I do see it, particularly in the early stages, as operating at least as much on a bilateral as on a multilateral level and the existence of confidence on both sides is absolutely essential. No individual would entrust his or her secrets to any other human being unless he or she was clear that those secrets would be kept. The same applies to intelligence agencies.
Your questions are particularly interesting because they are particularly difficult and because they address ultimately unsolvable problems, it does not mean that we cannot produce at least partial solutions. For example the need to strike a balance between the need for secrecy and the requirements of what you described as the information society, the public's right to know, these are exceptionally difficult issues. I think the answer is that when you have to balance two important priorities there is no fixed solution which is valid all the time. For example, how do you balance liberty and security. Now there is never a complete answer, both are important priorities but the balance between them changes according to circumstances, for example in war time and for that matter in major international crisis, security tends I think to be given a higher priority relatively to the liberty of citizens than in peace time. In peace time the freedom of individual citizens, which includes the freedom of information, is naturally given a higher priority in relation to the needs of national security than in war time. What I think we can say however is that this problem has only really been addressed in a serious way over about the last quarter of the century. In this period of time we have made considerable progress, but we are still learning. It is very difficult now to remember that in Britain, which had a very exaggerated notion of official secrecy, all Governments, up to and included that of Margaret Thatcher, in the 80s, agreed that we could not even officially admit that we had a foreign intelligence service. So it was not until the Queen's speech of 1992, that the British Government actually admitted for the first time that it had a foreign intelligence service, the Secret Intelligence Service (SIS), better known to the British media as MI6. Before 1992 every British Government said no to a Parliamentary Committee which had access to any part of secret intelligence. Well we now have, and we have had for eight years, the Intelligence and Security Committee, in other words probably the best compromise available. This Committee consists of Members of Parliament, members of the two Houses of Parliament, and produces an annual report on the performance of the intelligence community. One of the astonishing things about that is, I think, how little publicity it has got in Britain. The reason for it had so little publicity is that the Committee has almost never failed to agree with the Services, that is despite the fact that it has had a membership which include people who are certainly not passive in their attitude towards power. It has included Campbell Savers, who in the 80s was one of the main parliamentary critics of British intelligence. More recently, it included the person who most people believe will be British Prime Minister in 20 years, Yvette Cooper. In other words, it has proved possible in Britain, and not only in Britain, for a bipartisan committee, which has included traditional critics of the intelligence community as well as traditional supporters, to agree on an annual report and to agree also, in general terms, that the intelligence community is doing a good job. So what strikes me is the progress made in less than a decade, it is almost impossible to realise as I'm talking to you that 10 years ago the British Government refused to admit that we even had a foreign intelligence service and refused also to accept that there could possibly be a Parliamentary Committee which might be allowed access to intelligence material. So that's been the British approach and it is a good example of how Britain has been able to learn from the experience of other countries, not simply the United States. Still one can not ignore the American response to the intelligence scandals of the mid 70s and how they provided a model which has been widely imitated in other countries: the setting up in other words of the Congressional Committees on intelligence, that is one of the turning points, I think, in the relationship between democracy and intelligence and has represented a model imitated, with variations of course, by many other countries.

Your criticism of what you define as "the sound-bite and instant opinion" represents a crucial passage in your speech. This is a warning to avoid that decisions, even vital ones for the international balance, are made without a deep and thought-out evaluation, to be drawn in secrecy. How to strike a balance between the need for secrecy, typical of the intelligence work, and the requirements of the information society? And also, how to combine the need for a long-term perspective with the need for immediate answers (the culture of the instant opinion stemming from the typical shortcoming of last century's final years: working with a short-term perspective)?

Well, I think the first thing to do is to recognise the absurdity of the "culture of the sound-bite and of the instant opinion". This is a short term deviation and not a long term trend. I think one of the sad things about the present time is that plenty of people opt for "the sound-bite and instant opinion".
The second thing that I would say is that there is no contradiction whatever between long term analysis and short term intelligence reports. Currently intelligence reporting on information, whether classified or unclassified, is going to have sense only in a long term context. It means that if you do not understand the context you will never understand any detailed information which arises.
So, in other words, the success of current intelligence reporting is dependent on A) an understanding of the long term context by the analyst and B) I would also say, an understanding of the long term context by the policy maker. Let me give you an example: the great contradiction in the history of Soviet intelligence is that their collection of intelligence was frequently superb in Britain and in Italy during and immediately after the second world war, and the cold war. But the ability of the KGB to use the political intelligence it obtained was very limited because of the complete, or very substantial misunderstanding of the context of Western policy making by the Soviet leadership. Just two examples: during the Second World war, at the peak of Soviet intelligence collection in Britain, it was able to rely on a group that was called "the magnificent five" young graduates from my University, in Cambridge.
The KGB regarded this group of five agents as probably the best in its history. They were providing spectacularly good intelligence in the middle of the second world war from the major intelligence agencies and from the corridors of power. For two years in the middle of the second world war, they were regarded (we now know) as a British deception. Why? Because what they were saying about the way British policy was made and what British policy consisted of was so at variance with the understanding of British policy by the Kremlin, Stalin and the KGB leadership, that they couldn't believe it was true. Even after the Stalinist era, intelligence analysis in the Soviet Union was distorted and usually seriously distorted. Why? Because it was felt necessary to tell the leadership only things which would not offend it. And the best evidence we now have is that Michael Gorbaciov from the moment he came to power in 1995 was really serious about what he called "the new thinking about the West". KGB stations around the world are sent a directive which has as its title "On the impermissibility of factual distortions in the actual state of affairs". And what does that mean? It means that for the first time KGB stations were ordered to tell the leadership of KGB, and the leadership of KGB to tell the leadership of the country only what is going on even if it is information which they do not want to hear, so I would have said that intelligence analysis in the Soviet Union only began to function in the final years because at long last Gorbaciov had access to a general understanding of the West. The Soviet example demonstrates how current intelligence reporting will not achieve its aim unless it takes place within a broad understanding of the context not only by the intelligence agencies but also by the policy maker

In your paper you argue that a threat is always present and does not change substantially, what changes instead is the nature of its source. Having said that, a combination of fanaticism strengthened by power and weapons (today, even WMD) appears to be ever present, and is the element which today major world powers find most difficult to control (this is the main danger). Intelligence plays an important role in identifying future threats to be tackled either by preventive military actions or by diplomatic means aimed at removing the dangers uncovered by intelligence itself. In such a perspective how is this crucial role intelligence plays changing?

My basic point is that the general nature of the threats to national security does not change very much over time, but its specific nature does, in other words, what disrupted the security of the twentieth century was what has been called "fanaticism armed with power" in other words despots like Hitler, Stalin, Mao Zedong, Pol Pot. Fanatics, who were prepared to use extreme means to achieve their aims.
Now the difference in the 21st century is that there are no longer major powers ruled by fanatics willing to use any means available to achieve their aims, there are some medium size powers who look a little bit like that, North Korea has been one such powers (we don't know if it is still is, it may be changing, there is some optimism) and Saddam Hussein has been another example. What we face is a multiplicity of threats. Now, how to identify them? Well, as usual those who say that you can find out everything you need from open sources are wrong, those who say the only way you can do it is by secret sources are wrong, these are both extreme positions. The answer is the same as it has always been, the information which States require for their national security has to be sought along a broad spectrum which goes from open information at one end, the kind of information which you can see simply by opening newspapers to the other end, information from most secret sources, so secret that they are known only to a very limited number of people.
Let me apply the particular point that I put: the need always in intelligence to combine open sources with secret sources to identify the current threats to national security. One of the things that has been very little noticed is that many, if not all, of the diverse groups threatening national security have in common the fact that they are all "conspiracy theorists", just as so many of the despots of the twenty of century were. Stalin's mind was polluted by conspiracy theories, Hitler was obsessed with the conspiracy theory about the Jews. Now all the groups which have caused problems over the last decade: Aum Shirinkyo, Timothy McVeigh, just executed in the USA, Osama Bin Laden and Islamic fundamentalist terrorists have a view of the world distorted by terrible and illogical conspiracy theories. The World Wide Web, the Internet is a wonderful source for identifying conspiracy theorists. Almost all the conspiracy theorists in the world advertise their conspiracy theories on the World Wide Web. Of course, the majority of the conspiracy theorists are not prepared to use violence to achieve their aims, but I think a good preliminary way of identifying threats to national security from the enormous diversity of groups which in recent years have included Aum Shirinkyo and Osama Bin Laden (their addiction to conspiracy has very little in common) is the www. For example, there are groups which advertise on the web both their conspiracy theories and arms training for their members. You don't know, but you may have identified a very serious threat. When it comes to getting more information about them, technical intelligence, overhead surveillance is a lot less important than it was during the Cold War. Satellite surveillance of the Soviet nuclear strike force was absolutely essential during the Cold War, but it is not nearly as useful in tracking Osama Bin Laden. What I think we are moving back to is a return to the traditional priority of the penetration agent, that is in dealing with threats such as the ones I have outlined there is absolutely no substitute for the penetration agent. The experience of the 20th century I think shows that this can be done, it is sometimes very difficult. It is no secret, it has become a matter of public knowledge, that in Britain and USA, the Real IRA, that part of the republican movement which is still continuing terrorist attacks, has been heavily penetrated. That is I think the solution. That is linked, so far as "rogue regimes" are concerned, to encouraging defectors. Again I think that the experience of the Cold War shows that can be done, but it also shows that many opportunities have been missed.
We can do it a good deal better than we use to I'm aware that many defectors can be extremely difficult people to deal with, but they can also provide irreplaceable intelligence.

Prof. Andrew. you prompt Intelligence Services to understand their past, in order to avoid making the same mistakes which, because of rivalries or misunderstandings, had catastrophic effects such as the Japanese attack at Pearl Harbour. At the same time, you present us with an analysis of the difficult relationship between Services and policy-makers. In authoritarian regimes, Services do not carry out their task, which is to tell policy-makers even those unpleasant truths they do not want to hear. The revolution of the Gorbaciov era involved also a new direction to be followed by KGB foreign residents' reports, they were not modified any more to meet Moscow's wishes or requirements. For policy-makers in democratic countries is fundamental to maximise the structural advantages of Intelligence Services aware of their role. Can you just go back onto this statement which, during the conference, raised some doubts in other speakers believing that policy-makers are not really interested in learning the truth (it might be true also for Intelligence Services). Is it possible to be unaffected by politicians' will and to ignore their expectations?

The final question is on the problem of policy makers not being really interested. I think that the more we know the more we see that Western intelligence analysis during the Cold War was very much better than Soviet intelligence analysis. Precisely because it was able to operate within a democratic system. The more the Western policy makers imitated the Soviet model, that is to say behaved in an authoritarian way, believed that intelligence reports had nothing to tell them of any importance, the less successful they were at using it. In other words, when you're dealing with an authoritarian personality the problem is this, if the intelligence report agrees with what the authoritarian policy maker thinks than the policy maker says "that's of no interest I knew that already". If it disagrees with what the policy maker thinks then the policy maker says "that's just wrong and it is of no interest". But in the end, it is possible I think to learn from experience, very little research so far has been done on the way in which policy makers use and have used intelligence. We know far more about how intelligence agencies have collected and analysed intelligence than on how policy makers have used it. I think that the gradual increase and dissemination of research on the subject is something which will over time at least create a climate of opinion in which it becomes increasingly apparent to policy makers (of course not all will learn the lesson), that the successful use of intelligence depends heavily on them. That's a difficult lesson to learn but of course it has been difficult for policy makers throughout the centuries. My final example: within the medieval courts of Europe, these were authoritarian structures, one of the great problems was telling the ruler things which he did not want to hear. So, many of them had what was called the "royal fool". He was there to entertain the ruler, he had a kind of a licence to tell him uncomfortable truths which other people did not have. It's absolutely essential, I think, that intelligence agencies should, to some extent, in modern democratic structures, be able to acquire the privilege of the medieval royal fool. That is to say to tell policy makers what they do not want to hear. It's a good deal better for policy makers to hear what they do not want to hear from their intelligence chiefs than from the newspapers, because what it is told them in secret is a good deal less humiliating than having your opinions challenged in public. It does involve a change of culture and that is the biggest challenge for intelligence communities in the 21st century. It's not the collection of intelligence that is a challenge, it's not the analysis of intelligence that is a challenge, it's not the relationship with other intelligence communities that is a challenge, it is the relationship between intelligence agencies and policy maker. That's the only partly solved intelligence problem of our times.

One last question: over the 21st century how is the relationship between open and confidential sources going to change? Will it be better to acquire information giving priority to the quantity (open sources) or to the quality (confidential information of well tested reliability)?

My own view about this is that is logically a false dilemma. All States for the protection of their national security require information which comes from an extremely broad spectrum ranging from what is publicly available to what is most secret. The proportion between the two will vary according to the subject. For example: on the world economy my own view is that very little is available, which is really important, from secret sources which is not available from open sources. The more secret the opponent and that plainly applies to close societies such as Iraq, such as North Korea as well as to terrorist groups, which by their very nature operate in secret, than higher priority should be attached to secret sources. Now the dilemma to which I have already referred is how to reconcile the need to protect the secrecy of intelligence with the public's right to know. That's an extremely difficult problem and there's no permanent solution to it, it has continually to be readdressed. I think that I would simply say this: when there is an important intelligence source and an important intelligence method you have to protect that is not simply for the present but also for the future. So in the end if putting information in the public domain compromises an intelligence source or an intelligence method which is going to be of future importance, then that has to have priority. Just two very obvious examples: the most important intelligence source which Britain had between the wars was its ability to decipher Soviet telegrams. It was able to do that for ten years after the Russian revolution. It gave the secret away in 1927 and it was never able to do so for a very long time. During the Second World War, with the intelligence codenamed ULTRA, Britain and the Anglo-American were able to decrypt high grade enemy ciphers and this shortened the Second World War, we don't know how much it was shortened, but it only shortened it because it was kept secret.
Had that source being compromised in the way that our ability to decipher Soviet telegrams had been compromised in 1927. The Second World War would have gone on longer. The only reason the secret was so successfully kept was that Winston Churchill had been one of those responsible for giving away the secret in 1927. He had an ability which many policy makers do not have, which many human beings do not have, the ability to learn from his mistakes. So the enormous priority attached to keeping the ULTRA secret during the Second World War, was the result of the awareness that an error had been made.
The most successful businessman in the world today we all know his name, his name is Bill Gates. When he interviews people for jobs in Microsoft he is said to say the following: " I don't want to talk to you about your successes because if you hadn't had any successes you wouldn't have the possibility of a job with Microsoft. What I want to ask you about is what were your mistakes and what did you learn from them". That's the great challenge I think for both intelligence agencies and policy makers. British and Italian intelligence communities, British and Italian policy makers in their use of intelligence have had considerable successes. They had also, like the rest of humanity throughout history failures, we all had failures. The critical question is what were your failures, do realise what your failures were and begin to learn from them. The biggest failure in Western intelligence during the 20th century was in the relationship between intelligence and policy makers, that is going to be resolved only if it is recognised by both intelligence agencies and policy makers, and it is not until they can resolve it that they would be..... offered a job with Microsoft.