by Stephen Grey
Spying’s Limitations
No estimation of the value of spying makes sense without first considering its limitations. Modern spying, just like ancient spying, never offers unqualified benefits. It can easily go wrong and is not without distinct and costly trade-offs. These trade-offs are important, because without knowing in advance what will succeed or fail, the decision to use a spy must always be a risk calculation, weighing up the potential benefit of success with the potential fallout from failure. It is not enough to point to one great success and imagine this justifies everything that follows.
The first trade-off, to borrow from science, may be called the ‘observer effect’, which is the term used to explain that the act of observation alters the object under observation. Roughly applied to human espionage, it means that the act of spying cannot be neutral. At some point it involves taking actions, any one of which carries the risk of discovery that may induce a hostile and counterproductive reaction. For instance, the inducements offered in spying – such as paying agents large amounts of money – may not only be seen as evidence of hostile intent if discovered, but also incentivize agents to cause events that would never otherwise have happened, in other words to act as provocateurs.
One major advantage of signals intelligence over human intelligence is that the observer effect has tended to be much weaker. A signals intelligence satellite in orbit 22,000 miles into space could hoover up signals even from friendly states with almost zero chance of anyone working out who is being listened to. This calculation is altering, however. The diplomatic fallout from Edward Snowden’s revelations about whose phone calls the US was listening to, including the German chancellor’s and the UN secretary general’s, showed that signals intelligence is not without risk of blowback. The widespread use of strong encryption also alters the calculation, since formerly passive signals intelligence agencies may need to take active measures, such as burglary, to steal the passwords used by their targets.
Another trade-off is the ‘action effect’, by which I mean that the use of intelligence tends to undermine its collection. This is because, consciously or unconsciously, an enemy will begin to notice when his secrets are turned against him. To take an extreme example, if an agent passes on details of a terrorist’s murderous plot and that plot is defeated, the terrorist may then suspect the agent of betrayal, tell him no more secrets or even kill him. As the British Army demonstrated in its handling of the agent Steak Knife in the IRA, there are many clever ways to muddy the waters and misdirect suspicions about who leaked information. But it cannot always be done. And even when no one knows who the traitor is, over the long term, by an evolutionary process, those who are more security-conscious and do not leak secrets to the agent are likely to rise in importance. The result of all this is that secret services, even when they have very good agents and good information, tend to be very cautious about encouraging anyone to make use of that information.
The third major trade-off could be called the ‘rogue effect’, which is the tendency of secret operators to go off the rails, the risk being that spying’s intrinsic secrecy divorces those involved from the norms of society. They lack the usual means of self-regulation in public life, notably judgement and scrutiny by the public.
In order to protect their tactics and the identity of their sources, spymasters remain cloistered in a kind of private club, in an isolated environment that can, without care, lead to rogue behaviour. Basic assumptions within this club can lie unchallenged, as can the veracity of their agents’ reports. Apparently, if their activities are kept secret, usually ordinary and decent people will do irrational and indecent things. Or as one British intelligence officer put it crudely, ‘Intelligence agencies whose operations are pursued without strict outside scrutiny invariably f*** up in the end.’
An example of this rogue behaviour was exhibited by the leadership of the CIA after the 9/11 attacks. They might have had their sign-off from the president and indeed reflected the vindictive public mood, but in countenancing systematic torture and a chain of secret detention places, they strayed far from the broader values of their society, or even the law. They had failed what should be called the ‘flap test’: that is, would a secret action be judged publicly acceptable if it were no longer secret? A less dramatic, but equally clear, example of aberrant behaviour was the Special Demonstration Squad (SDS) in London’s Metropolitan Police, which, over the course of four decades, believed it acceptable – in the name of quelling protests by environmental activists, for example – for their agents to sleep with their surveillance targets and even father children (some of whom, it is alleged, they abandoned).
This was a genuine rogue unit. An investigation by the Guardian found that of nine undercover policemen identified, ‘eight are believed to have slept with the people they were spying on’.10 But, when ten women sued the Metropolitan Police, claiming they had been deceived, a judge came to the conclusion that so-called sexspionage was not unusual. Mr Justice Tugendhat, a High Court judge, said examples came to mind from the realms of fiction.
James Bond is the most famous fictional example of a member of the intelligence services who used relationships with women to obtain information, or access to persons or property. Since he was writing a light entertainment, Ian Fleming did not dwell on the extent to which his hero used deception, still less upon the psychological harm he might have done to the women concerned. But fictional accounts (and there are others) lend credence to the view that the intelligence and police services have for many years deployed both men and women officers to form personal relationships of an intimate sexual nature (whether or not they were physical relationships) in order to obtain information or access.11
How far should a spy go? It was an open question. But certainly not that far to deal with such a small threat. The SDS had been formed largely of uniformed policemen with training as neither detectives nor undercover agents, in stark contrast to Scotland Yard’s professional undercover unit, for a long time designated SO10. ‘It isn’t normal to sleep with a target. If you have to, it means you are not in control,’ said one former operative.
The Misuse of Spies
Pure spying, then, has many weaknesses, tending to undermine its value. But the biggest drawback of all comes not from intelligence collection itself but rather from the temptation to intervene and misuse that intelligence too readily. Modern society has developed great techniques to pry into the lives of others – the challenge is how we make use of these.
Society has faced these dilemmas before, but they were of a different character. In the Cold War, the issue was civil liberties. A substantial amount of spying was done by East and West against their own citizens, with the object of preventing subversion. But this information collected secretly was also used by the state to take pre-emptive action. So, for example, the careers of East Germans discovered to have contacts with the West were secretly hindered. And in the West, those with suspected communist sympathies were secretly blacklisted from taking up certain jobs and radical organizations were secretly subverted if they were seen as communist fronts. In essence, this spying was objectionable because it was an affront to natural justice and an open society in which someone’s faults or blessings could be debated openly and fairly.
In the twenty-first century, the threat to civil liberties continues, even if it has altered. Intelligence collection is – contrary to some reports – far more tightly focused on those who are suspected of posing a violent threat to society and rarely directed against domestic ‘subversives’. But when violent threats are identified, political leaders continue to look for a convenient and secret response. If a group of Britons in Pakistan are heard discussing bombing a shopping mall in New York, it might be tempting to think a convenient explosive dropped from a drone would deal with the problem. Or if, as in Britain, assassination is ruled out but the only evidence of the plot is secret intelligence, then it might be tempting to lock the plotters in jail using secret court procedures. As be
fore, this poses a threat to natural justice. Supposing the intelligence is wrong? Is this action fair and proportionate?
But taking preventive action in this way also enlarges the role of secret services, moving intelligence into the uncomfortable and rarely accurate world of prediction. How often do people plan to commit a crime that may never come off or that they may, in the end, decide not to commit? We have found very powerful ways to reach into people’s thoughts. The dilemma for society is when it is right to intervene and punish those intentions.
Judging When Spies Are Effective
While, as we have discussed, spying today has its limits and can be abused, one reason it persists is that mechanisms have been put in place to compensate for those weaknesses. A tendency to go rogue, for example, is prevented by strict political accountability.
Four observations on the effectiveness of spying were outlined earlier, drawn from experience in the Cold War: that activity is not the same as achievement; that human intelligence offers the most when it is corroborated or, better still, verified; that spying proves itself valuable when it is highly focused and politically directed; and that spying has to be a weapon of last resort. These principles apply equally to modern spying.
First, as before, the mere existence of a spy in the enemy camp is not sufficient to be of value. Modern technology and techniques can make spying more efficient than it has been before. Some spy missions are really successful and make a genuine difference in changing government policy or averting some crisis or crime, as, say, when the UK’s agent in Yemen prevented an attack on an airline in 2012. But these successes are rare, even if, given the scale of potential security threats, it seems worthwhile to persist.
Second, it also remains true that many of spying’s limitations – the potential weakness of information provided by a serial betrayer, the risk of a source being exposed and the effects of acting on intelligence – are all less worrisome when that intelligence can be corroborated. One former intelligence officer described how in Afghanistan, for instance, the military wisely ignored reports from secret agents that a Taliban group were in this village or that village, but when that report was corroborated by signal intelligence – for instance, by locating the mobile phones of identified Taliban fighters in that village – then they were willing to commit forces to attack the group. The HUMINT could not be trusted alone, but it helped to narrow the target for surveillance and thus, in the end, to locate the enemy with reasonable certainty.
Third, political direction and a tight focus remain key to successful HUMINT. As neither Sunni extremism nor Iraq’s weapons programme was a focus in the early to mid-1990s, policymakers paid the price later when they found they had no spies in place where they needed them. The US, with its aim to remain a superpower and have global influence, has a particular problem with such focus. It tries to collect intelligence from too many places and, despite its huge resources, tends to underperform. But with strong political direction and by concentrating spying resources on key threats, agencies have a chance to make the recruitments they need.
Political direction also means political accountability. In most democracies, to guard against the tendency to go rogue, secret services require political approval for their operations. In the US, most covert actions are signed off by the White House, if not the president himself. In Britain, all non-trivial actions by SIS, anything likely to have repercussions, are signed off by the foreign secretary. And the system does work, even if successive scandals appear to show these mechanisms are still far too weak.
Finally, spying remains useful and successful when used as a last resort. It is a hostile act: always invaluable during warfare, but often counterproductive and always to be used sparingly in peacetime.
* * *
Spying on the enemy has been seen as crucial to military strategy on the battlefield since ancient times, with the resulting intelligence being used for tactics of surprise and ruse. ‘All warfare is based on deception,’ wrote Sun Tzu in 400 BC. ‘There is no place where espionage is not used.’12 But in modern warfare, intelligence is even more valued, particularly as a means of replacing the grinding war of attrition – as seen in the trenches of the First World War – with victory based on the concentration of overwhelming force on the enemy’s weak spots. Such an approach depends on both mobility and good information about the enemy and their plans.
Whether it is war in ancient China or against al-Qaeda, running human spies is just one way of filling in the canvas of a broad intelligence picture. But war fundamentally changes the risk calculation in espionage. In the Second World War, an agent who parachuted behind enemy lines faced a very high chance of capture or death, or both. But when good intelligence might save hundreds of lives, as, for example, in the D-day landings in Normandy, the risks to the agent were worth taking.
Because the Cold War was ‘cold’, the risk of death was usually remote. The Soviet Union executed its traitors, and so the West’s recruited agents always risked their lives. But the intelligence officers who handled these agents were much safer. By tacit agreement, the superpowers never tried to assassinate each other or take reprisals. These safety guarantees, however, were of little value when the West got embroiled in ‘hot’ conflicts in the developing world. The war in Vietnam in the 1960s and 1970s brought mortal danger for US field operatives, as did the civil war in Lebanon in the 1980s. In each conflict, both CIA officers and agents were killed. The bombing of the US Embassy in Beirut in 1983 caused what is still the biggest CIA loss of life, with eight officers killed. In recent years, the War on Terror and the conflicts in Iraq and Afghanistan have brought new dangers.
Britain’s SIS has always operated with great caution. No British officers have lost their lives in action since the Second World War. However, according to insiders, the lives of a series of agents working for the British have been lost while infiltrating Islamic militant groups. Given al-Qaeda’s callous willingness to shed innocent blood, though, spying on them was, in principle, worth that sacrifice.
In peacetime, spying has had a very mixed record. In the fight against crime or domestic extremism, it coexists poorly with a criminal justice system that guarantees a fair and open trial, and it is hard to stop agents provoking crime. It should be used, but sparingly and only with great expertise. It may, for example, be essential as a means of tracing the hidden hand of the powerful gangsters who are the instigators and beneficiaries of serious crime.
Between peaceful nations spying is automatically dubious. The unmasking of a traitor or any attempt to recruit a spy tends to sow enmity. Where secret agencies have survived and grown in peacetime, it is because the fear of war is ever present. In particular, since the US detonated the atom bomb over Japan in 1945, the fear of nuclear conflict has made their existence, particularly in nations that possess nuclear weapons, hard to contest. Spying might often be expensive, inefficient and ineffective, but not always. Set against the prospects of nuclear war, that distinction makes all the difference. However much politicians might deride the titbits that their spies serve up, none of them could risk not having an intelligence service.
From our look at the Cold War, it is clear – with hindsight – that spying could raise the tension at times but also lower it, helping deal with paranoia about the other side’s intentions. And if a scorecard existed, it would have to record that both the KGB and the CIA proved their ability to steal military secrets and to develop great systems of early warning. Military spying evened out the contest; early warning helped calm the nerves. And together they did, in their small and highly expensive way, help to keep the peace. Neither agency was ever a great success at political spying in the other’s homeland. The West never noticed that the East was collapsing and the East never realized that it was failing. Up to a point, this same story of tactical brilliance and strategic myopia has continued.
* * *
In the new world of espionage, despite initial doubts after the fall of the Berlin Wall, the secret se
rvices won a partial reprieve by shifting the case for spying into a struggle against other non-state threats. The threats were not nuclear holocaust but dirty bombs; not invasion by Warsaw Pact troops but a massacre of the innocents, whether by warring tribes or a bomb in a shopping centre. Though less of a threat to the state, the damage they could cause was real and tangible to the public, and so the need for intelligence – and for spies – was arguably unassailable. And in declaring War on Drugs and then War on Terror, US politicians were beginning conflicts that might conceivably never end.
Whether or not combating terrorism is really a ‘war’, it still demands a proportionate response, but one in which spying remains a powerful weapon of last resort. As John MacGaffin, a thoughtful former senior CIA operations officer, wrote, ‘Clandestine collection of HUMINT must be employed only in pursuit of information that is truly essential to the most critical tasks of civilian and military national security affairs and only when that information cannot be acquired in any other way. When either of these two conditions is missing, the outcome almost always suffers.’13
The results of spying first creak and then fall apart when they are relied upon too heavily and become the sole means of supporting critical actions. The use of military and intelligence methods to round up al-Qaeda prisoners and send them to Guantanamo Bay after 9/11, for example, may have seemed to make sense at the time. But it was counterproductive in the long term. The evidence collected against these men was usually secret intelligence and so was no use in a court of law. That has made it very hard for the US to decide what to do with them, and to justify their continued, indefinite detention.
