The inability of the United States government to anticipate the ISIS offensive that has succeeded in taking control of a large part of Iraq is already being referred to as an “intelligence failure.” To be sure, Washington has unparalleled technical capabilities to track money movements and to obtain information from the airwaves. It is adept at employing surveillance drones and other highly classified intrusive electronic methods, but there is an inherent problem with that kind of information collection: knowing how the process works in even the most general way can make it relatively easy to counter by an opponent who can go low tech.

Terrorists now know that using cell phones is dangerous, that transferring money using commercial accounts can be detected, that moving around when a drone is overhead can be fatal, and that communicating by computer is likely to be intercepted and exposed even when encrypted. So they rely on couriers to communicate and move money while also avoiding the use of the vulnerable technologies whenever they can, sometimes using public phones and computers only when they are many miles away from their operational locations, and changing addresses, SIM cards, and telephone numbers frequently to confuse the monitoring.

Technical intelligence has another limitation: while it is excellent on picking up bits and pieces and using sophisticated computers to work through the bulk collection of chatter, it is largely unable to learn the intentions of terrorist groups and leaders. To do that you need spies, ideally someone who is placed in the inner circle of an organization and who is therefore privy to decision making.

Since 9/11 U.S. intelligence has had a poor record in recruiting agents to run inside terrorist organizations—or even less toxic groups that are similarly structured—in places like Afghanistan, Iraq, and Syria. Information collected relating to the internal workings of al-Qaeda, the Taliban, dissident Sunni groups in Iraq, and now ISIS has been, to say the least, disappointing. To be fair this is often because security concerns limit the ability of American case officers to operate in areas that are considered too dangerous, which is generally speaking where the terrorist targets are actually located. Also, hostile groups frequently run their operations through franchise arrangements where much of the decision making is both local and funded without large cash transfers from a central organization, making the activity hard to detect.

In the case of ISIS, even the number of its adherents is something of a guesstimate, though a figure of 5,000 fighters might not be too far off the mark. Those supporters are likely a mixed bag, some motivated to various degrees by the ISIS core agenda to destroy the Syrian and Iraqi governments in order to introduce Sharia law and recreate the Caliphate, while others might well be along for the ride. Some clearly are psychological outsiders who are driven by the prospect of being on a winning team. They are in any event normally scattered over a large geographical area and divided into cells that have little in the way of lateral connection. They would, however, be responsive to operational demands made by the leadership, headed by Iraqi Abu Bakr al-Baghdadi. Moving in small groups while lacking a huge baggage train or infrastructure, it was relatively easy to concentrate to push into Iraq and link up with dissident Sunni tribesmen without necessarily coming to the attention of spies in the sky, American drones flying out of Turkey.

It should be assumed that the U.S. intelligence community has no spies inside ISIS at any level where it might be possible to collect significant or actionable information. In the past, successful penetration of a terrorist organization has come about when a dissident member of the group surfaces and volunteers his services in return for money or other considerations. This is how law enforcement and intelligence agencies broke the Euro terrorists who were active in the 1970s and 1980s, but its success depended on the radical groups being composed largely of middle-class students who were ideologically driven but by nature not necessarily loyal to a political cause or its leaders. The defector model does not appear to have been repeated successfully recently with the demographically quite different radical groups active in Syria, at least not at a level where actionable intelligence might be produced.

Lacking a volunteer, the alternative would be to run what is referred to as a seeding operation. Given U.S. intelligence’s probable limited physical access to any actual terrorist groups operating in Syria or Iraq any direct attempt to penetrate the organization through placing a source inside would be difficult in the extreme. Such efforts would most likely be dependent on the assistance of friendly intelligence services in Turkey or Jordan.

Both Turkey and Jordan have reported that terrorists have entered their countries by concealing themselves in the large numbers of refugees that the conflict in Syria has produced, and both are concerned as they understand full well that groups like ISIS will be targeting them next. Some of the infiltrating adherents to radical groups have certainly been identified and detained by the respective intelligence services of those two countries, and undoubtedly efforts have been made to “turn” some of those in custody to send them back into Syria (and more recently Iraq) to report on what is taking place. Depending on what arrangements might have been made to coordinate the operations, the “take” might well be shared with the United States and other friendly governments.

But seeding is very much hit or miss, as someone who has been out of the loop of his organization might have difficulty working his way back in. He will almost certainly be regarded with some suspicion by his peers and would be searched and watched after his return, meaning that he could not take back with him any sophisticated communications devices no matter how cleverly they are concealed. This would make communicating any information obtained back to one’s case officers in Jordan or Turkey difficult or even impossible.

All of the above is meant to suggest that intelligence agencies that were created to oppose and penetrate other nation-state adversaries are not necessarily well equipped to go after terrorists, particularly when those groups are ethnically cohesive or recruited through family and tribal vetting, and able to operate in a low-tech fashion to negate the advantages that advanced technologies provide. Claiming intelligence failure has a certain appeal given the $80 billion dollars that is spent annually to keep the government informed, but it must also be observed that it is also a convenient club for Republicans to use to beat on the president, which might indeed be the prime motivation.

The real problem for Washington is that penetrating second-generation terrorist groups such as those operating today is extremely difficult, and is not merely a matter of throwing more money and resources into the hopper, which has become the U.S. government response of choice when confronted by a problem. Success against terrorists will require working against them at their own level, down in the trenches where they recruit and train their cadres. It will necessitate a whole new way of thinking about the target and how to go after it, and will inevitably result in the deaths of many more American case officers as they will be exposed without elaborate security networks if they are doing their jobs the right way. It is quite likely that this is a price that the U.S. government will ultimately be unwilling to pay, and that unreasonable expectations from Congress will only result in more claims that there have been yet more intelligence failures.

Philip Giraldi, a former CIA officer, is executive director of the Council for the National Interest.