The end of the cold war must not lead to the end of spying, argues Christopher Andrew, because there are plenty of deviant dictators and fanatical terrorists we need to keep an eye on.

It is much argued these days that now that the Soviet Union is no more, there is no need for post-cold war intelligence services either. The cold war, it was generally accepted, made secret intelligence necessary. But Theodore Draper is only one of a number of prominent American critics of the post-cold war intelligence community who has recently dismissed the CIA as an irrelevant "relic that the country does not need".

As with other policy debates on both sides of the Atlantic, I believe that public discussion of the future role of the secret services has been degraded by short-termism. The need for a 21st century intelligence community derives not from the second world war, the cold war, the end of the cold war, or any other transient phase of human history, but from the long-term development of international relations between states. For the foreseeable future we shall be faced with threats to national and international security. The effect of globalisation has been to make these threats more varied than during the cold war. In the global village what happens on the other side of the world affects us more directly than ever. And in order to monitor these changing threats, governments will require access to secret as well as to open information, such as that from diplomatic or media reporting.

Anyone who disputes the case for secret intelligence has to argue one of two deeply implausible propositions: either that there are no longer any significant threats to national security: or that there are such threats, but that they can be monitored via open sources.

É«ºÐÖ±²¥

It is true that identifying future intelligence priorities is difficult. A sane security system to identify the kind of threats governments need to monitor requires two strategic guidelines. The first is to accept the limitation of prediction. A merciful providence does not intend us to see more than the general outline of the problems that await us. We can see future threats to our security only as St Paul glimpsed heaven, "through a glass, darkly". The second essential is to base the assessment of threat on long-term trends, not on short-term fluctuations, in the international system.

All human inventions have sooner or later spread around the world. Thanks to globalisation, they now do so sooner rather than later. It is idle to suppose that weapons of mass destruction - chemical, biological and nuclear - will prove the exception to this iron law of history. The main intelligence priority of the new millennium will be to monitor, slow down, and, when possible, frustrate this ultimately unstoppable trend.

É«ºÐÖ±²¥

The other great constant of human history is mankind itself. Human nature appears to have changed little - if at all - during the approximately 30,000 years since Homo sapiens sapiens succeeded the Neanderthals. It is most unlikely to evolve significantly in the foreseeable future. So the international system of the next century, like that of the 20th, will contain both national leaders of sense and moderation, and a minority of deviants bent on wreaking havoc. Some of the deviants will have, and be willing to use, weapons of mass destruction. Secret sources of information will be needed to monitor the deviants.

The spread of weapons of mass destruction will threaten both European security and the world order in the early 21st century. There will be crises in which rogue regimes threaten the West; conflicts in the rest of the world in which the combatants seem likely to use such weapons against each other; and the use of the weapons by terrorists. These are not theoretical dangers. The first examples of all three have occurred. Since intelligence, or the lack of it, has been critical in dealing with these incidents, it is reasonable to assume that it will continue to be.

The non-European power which has come closest to challenging the West is Iraq. The Israeli attack on Saddam Hussein's nuclear reactor in Osiraq in 1981, prompted by intelligence reports, had an importance which few suspected at the time. But for that attack, the odds are that Saddam would have entered the Gulf war with a nuclear arsenal. Saddam did, however, begin the war with large stocks of sarin nerve-gas warheads, and nearly placed them on his Scud missiles. He has been determined since his defeat both to continue and to conceal his biological weapons programme.

Probably the most dangerous rogue regime is Iran. According to probably reliable intelligence reports, President Rafsanjani boasted in 1995 that Iran had completed the first stage in the production of a nuclear bomb. Both the CIA and Israeli intelligence calculate that, within 18 months, Iran will have nuclear missiles with a range long enough to reach Israel.

The first crisis in the third world in which combatants seemed likely to use weapons of mass destruction, little noticed by the media, occurred in the Indian subcontinent in the spring of 1990. India massed 200,000 troops in the disputed terrotory of Kashmir, close to Pakistan. In a conventional war, Pakistan would have risked a repetition of its disastrous defeat by India in 1971. Intelligence reports to the then US president, George Bush concluded that Pakistan had at least six nuclear weapons, and might decide on a strike against New Delhi rather than run the risk of another humiliation. India, with a larger nuclear arsenal than Pakistan, would certainly respond in kind.

É«ºÐÖ±²¥

"The intelligence community", recalls Bush's adviser Robert Gates, "was not predicting an immediate nuclear war, But it was predicting a series of clashes that would lead to a conventional war that it believed would then inevitably go nuclear". In mid-May 1990 Bush sent Gates to Islamabad and New Delhi, to appeal for restraint from both sides. Whether or not there was a serious risk of the use of nuclear weapons in May 1990, there is little doubt that this crisis foreshadows others. Any renewal of the Bosnian conflict, for example, would pose a serious risk. While peace talks were underway at Dayton, Ohio, in 1995, UN forces discovered the remains of a Serb factory near Mostar designed to manufacture the fatal gas, sarin. The use of sarin on the Tokyo underground in 1995 by the Japanese cult, Aum Shinrikyo, may have been the first example of terrorist use of weapons of mass destruction, but it will not be the last. Had Aum Shinrikyo succeeded in covering its tracks in Tokyo, it planned to release ten more tonnes of sarin on the Washington and New York subways. There is no substitute for penetration agents ("moles") in keeping track of Aum Shinrikyo's successors.

Until quite recently, terrorists caused relatively few deaths. During the century before the Oklahoma City bombing, there were fewer than a dozen terrorist attacks which caused as many as 100 deaths. Traditional terrorists, such as the IRA, though happy to kill handfuls of people, have usually been more anxious to cause panic and publicise their cause than to instigate large-scale massacres.

All that is changing. Today's religious and cult-based terrorists are much more menacing than their predecessors. The most dangerous delude themselves into believing that they are doing the will of God in destroying the forces of Satan. According to Yoshihiro Inoue, the intelligence chief of Aum Shinrikyo, "We regarded the world outside as evil, and destroying the evil as salvation." This new generation of terrorists has more in common with the bloodthirsty fanatics of the age of religious warfare than with most of their 20th-century predecessors. It is probably just as well that Stephen Landor, the director-general of MI5, is the first British intelligence chief with a PhD in 16th-century religious history.

É«ºÐÖ±²¥

Thirty years ago there was not a single religious or cult-based terrorist group anywhere. As recently as 1980 only two of the world's 64 known terrorist groups were religious. Since then, however, Shi'a extremists alone have probably been responsible for more than a quarter of the deaths from terrorism. Sheikh Omar Abdel Rahman, the cleric found guilty of inspiring the World Trade Center bombing, told his followers, "We have to be terrorists... The Great Allah said, 'Make ready your strength I to strike terror into the enemies of Allah'." Terrorists deranged enough to believe that they are doing the will of God are also liable to believe they have divine authority to massacre as many victims as they wish.

The coming crises of weapons proliferation, though not the only reasons why intelligence agencies remain necessary, are the most predictable of the problems they face. Public opinion, however, does not take seriously this threat nor is it likely to do so until there is a catastrophe. There are no votes to be won in Europe by devising counter-proliferation measures. The situation resembles that just over a decade ago, before the 1986 Chernobyl disaster in the Soviet Union. Western experts knew that some Soviet-bloc nuclear reactors were unsafe and that there was a serious risk of disaster. Before Chernobyl, however, their warnings had little impact on western governments and even less on western public opinion.

The gap between public apathy and the reality of proliferation is worrying. So long as that gap remains, the argument that intelligence agencies are irrelevant relics of the cold war will continue to find sympathy.

Christopher Andrew is professor of modern and contemporary history, Cambridge University.

É«ºÐÖ±²¥