In the war between humans and machines, the humans have found a new way to strike back.
I'm not talking about a science-fiction shootout between Hollywood heroes and rogue cyborgs, like TheTerminator. I'm talking about the real-life battle going on today in Pakistan and Afghanistan. The unmanned aerial vehicles engaged in this battle—drones—aren't alien or autonomous. They're built, deployed, and controlled by the United States. In the last year, they've hit al-Qaida and the Taliban with more than 50 fatal missile strikes. By some accounts, the militants are so rattled that they're abandoning the mountains and moving to Pakistani cities, hoping the drones won't dare to strike there.
How can our enemies fight back? By targeting the machines' weak link: us.
The drones depend on human input. They need human authorization to fire. And to find enemy honchos and hideouts, they need targeting intelligence from human informants on the ground. Two years ago, the insurgents took aim at both of these inputs. They accused local people of scouting targets for the drones and butchered them in public to deter such spying. They also set off bombs in Pakistan to intimidate the Pakistani government into demanding an end to the drone strikes. But Pakistan didn't buckle, and the drone strikes have increased in tempo and precision. Apparently, the drones' managers have found plenty of new spies to replace the dead ones.
Now the Taliban seem to have come up with a new strategy: using the drones' human intelligence networks to infiltrate the program and kill the people who run it.
Last week, a suicide bomber blew up seven CIA officers at a U.S. military base in Afghanistan near the Pakistan border. It looked like just another insurgent attack. But it was more than that. In separate interviews, representatives of two Taliban factions have claimed that the mission's target was the drone program. "We attacked this base because the team there was organizing drone strikes," a commander allied with the Afghan Taliban told the Wall Street Journal. He said the attack was timed to kill the woman who led the team, since the Taliban knew she would be there that day. A Pakistani Taliban commander told the AP a similar story and added that the bomber was recruited as a "CIA agent" but turned against the agency.
Well, the Taliban say a lot of things. But in interviews with the Journal, the New York Times, and the Washington Post, current and former U.S. officers confirm the main elements of the story. U.S. personnel at the Afghan base were closely engaged in selecting drone targets in Pakistan. And they did this job, in part, by recruiting and interviewing informants on site. That would explain why the bomber targeted them and how he got in.
According to at least three reports, the bomber was recruited as an informant, invited to the base, and allowed past an initial checkpoint without being searched. Why wasn't he searched? One reason, the Journal reports, is that the base's CIA officers limit searches of such recruits "in the hopes of establishing trust." Another reason is speed. A former intelligence officer calls the incident an "asset meeting gone bad" and explains the lax search protocol as part of the CIA officers' strategy: "They felt the need to gather viable, time-sensitive intelligence was so pressing that it justified the trade-off."
Time-sensitive intelligence. That's the key phrase. The reason you don't make people go through a lot of screening to get into your facility is that their information might be hot. Who needs such hot information? The drones. Two minutes' delay can cost them a clean shot at a Taliban or al-Qaida commander.