On May 26, 2012, a missile from a drone hit the top floor of a two-story bakery in Pakistan’s North Waziristan tribal agency, killing four al-Qaeda members. A couple of weeks later, a similar strike took place when a drone launched two missiles at the upper level of a grocery store in Miramshah Bazaar, killing another senior al-Qaeda commander. The buildings remained intact, and no civilians were killed.
Sound familiar? On July 31, 2022, al-Qaeda chief Ayman al-Zawahiri was standing on the balcony of a safe house in Afghanistan’s capital, Kabul, when two Hellfire missiles fired from an unmanned aircraft killed him. No civilians were reported killed, while unconfirmed photos of the drone strike showed only partial damage to the balcony of the house he was apparently living in.
Although 10 years apart, the drone strikes are remarkably similar. Indeed, the drone strike in Kabul’s affluent downtown neighborhood eliminated a very high-value target in comparison to the North Waziristan bazaar strikes, which targeted makeshift improvised explosive device factories. Nevertheless, the quality of the intelligence locating and identifying the targets, the decision to conduct the attack at an opportune time, and the ability to avoid civilian casualties and widespread property damage in a closely-built area are very similar. This shows that drone strikes can be precise and discriminating and achieve counter-terrorism goals.
The attack on al-Zawahiri after America’s troop withdrawal from Afghanistan is exactly what President Biden’s “over-the-horizon” counterterrorism strategy sought to deliver. It was a model strike. But as critics have pointed out, it is just one data point.
However, we have more data showing how effective this strategy can be. The drone program in Pakistan, particularly under Obama’s second presidential administration, targeted and killed top al-Qaeda commanders as well as chiefs and militants of both the Afghan and Pakistan Taliban with minimal civilian harm. From 2009 onwards, even though drone strikes increased in number, civilian casualties decreased with every subsequent year. From 2013 onwards, no local civilian was reported killed by American drone strikes in Pakistan.
My own quantitative with the qualitative research confirms that over-the-horizon strikes can be carried out consistently and effectively with minimum casualties or popular blowback. I collected data by interviewing 116 residents of Pakistan’s tribal belt — where the United States has conducted over 420 drone strikes. Two-thirds of my sample said that they were in favor of the American drone program, calling it “successful” and “effective” in killing militants. All those in favor of unmanned strikes admitted that there were instances of civilian casualties, such as the attack on a Taliban commander’s funeral in 2009 and another strike at a jirga in 2011. However, the respondents quickly added that a majority of the strikes killed only militants.
Before beginning my research, I believed that drone strikes killed more civilians than militants. Media coverage of American drone strikes is dominated by images of houses reduced to rubble, news articles about scores of estimated deaths of women and children, and videos of charged Pakistani protestors setting American flags on fire.
But as I spoke to more and more local residents from Pakistan’s tribal areas — tribal leaders (known as malik) and elders (known as masharaan), university students, journalists, academics, activists, lawyers, and internally displaced persons returning to their homes — a clear majority of them rejected that narrative. For them, drones had done what local militias and even Pakistani military operations could not do: kill the bad guys, with limited-to-no collateral damage. “The drone is a justice-delivering technology,” one respondent from South Waziristan said. “Taliban Inc. — which is very powerful as it’s challenging Pakistan’s state sovereignty — commits atrocities against the local population and the people don’t have the power or the resources to retaliate or defend themselves. The drone compensates for this asymmetry.”
In the course of my research, I spoke to residents from all seven tribal agencies and frontier regions, seeking out the most representative sample of people possible from different socio-economic and political backgrounds. Among lower-income families or supporters of more overtly religious political parties, there was some sympathy for al-Qaeda and related groups. But the Pakistani Taliban’s violence and brutal imposition of its own version of sharia had created a widespread animosity that transcended many tribal, economic, and political divides. And this translated into support for drone strikes.
A resident of North Waziristan said he had seen drones hovering in the sky every day for weeks on end but had never witnessed a strike. “We never had a drone strike in our village because we didn’t have any militants in our area,” he said, demonstrating his confidence in the targeting of the drone program.
To be sure, this was not always the case. According to my respondents, the initial years of the drone program under President George W. Bush invoked fear among the population — missed targets caused trauma, and there were tens of civilian casualties, including women and children. Residents in my sample who support drone strikes had opposed the program during its initial years, but between 2009 and 2011, their opinion became more favorable and accepting of the operations.
The residents of Pakistan’s tribal areas began to support the drone operations because the unmanned program evolved and improved its targeting practices. During Obama’s first term, improvements in technology and intelligence collection and analysis made the targeting more precise and discriminating. The United States expanded and improved its intelligence-surveillance-reconnaissance capabilities by procuring more efficient drones, employing more analysts and linguists, and developing its sprawling intelligence-collection and data-fusion infrastructure. During this time, the Pentagon maximized the production and deployment of the Reaper — the Predator drone’s more lethal and efficient cousin — while at the same time ramping up training for operators for these new drones. This improved intelligence apparatus allowed the United States to locate militant networks and infrastructure and identify them correctly, helping to achieve precision: The intended target was accurately attacked.
Precision was not enough though — strikes had to ensure no civilians were killed. In 2011, a drone launched two missiles in quick succession at a jirga in Datta Khel, North Waziristan, killing over 40 local tribesman and Khasadar (members of a lightly armed local police force). The attack on a peaceful jirga led to disagreements on targeting practices and procedures within the Obama administration, resulting in an internal review. The review created more checks and balances in the selection and execution of a target: The State Department now had more of a say in the strike decisions, Pakistani leaders would get advance notice about some drone strikes, and the CIA would halt unmanned operations when Pakistani officials visited the United States. These changes had an impact, as demonstrated by the steady decline in civilian casualties.
By 2013, the Obama administration was feeling rising global pressure to ensure greater transparency and accountability regarding his classified unmanned program. As drone warfare was here to stay, particularly in areas of active hostilities that were not declared warzones, Obama introduced policy guidance on drone strikes and established an inter-agency review process to systemize and bureaucratize the process of target selection and authorization.
He promulgated the Presidential Policy Guide, which established the “near-certainty standard” of no civilian casualties, emphasizing that U.S. government agencies must identify targets for direct lethal action and also present a legal and threat-based justification for their inclusion. From 2013 onwards, not a single local civilian death was reported in American drone attacks in Pakistan’s tribal areas.
Weakening Terrorist Groups
Scholars have provided evidence of how the U.S. drone program in Pakistan and Afghanistan affected al-Qaeda’s organizational capabilities. Data on the Pakistani Taliban shows that its top commanders, including all its former chiefs, have been killed by drone strikes: Baitullah Mehsud, Hakimullah Mehsud, and Maulana Fazlullah were targeted and killed in 2009, 2013, and 2018, respectively.
American drone operations forced the leadership of the Pakistani Taliban (also known as the Tehreek-e-Taliban Pakistan) into hiding. There was no specific time when the drones would surveil or strike — they would hover or attack during the day and well into the night. “The only certainty was that it was safe for the Taliban when it was cloudy; no drones could be in the sky when it was covered with clouds,” explained a North Waziristan resident. Under these circumstances, all the Pakistani Taliban leaders could not be in the same place and at the same time, which the leaders found deeply frustrating. This restricted mobility and limited communication, which further weakened their command and control.
Moreover, the accurate targeting of top leaders and commanders created multiple leadership crises within the Pakistani Taliban, which in turn led to fragmentation and infighting. The infighting took place usually over disagreements about succession, with different tribes and factions wanting their commander to take the helm of affairs. This resulted in violent clashes which deepened cleavages, leading to fragmentation and desertions, further eroding their capacity to organize and operate.
All this caused the Pakistani Taliban to eventually lose their grip over the country’s tribal areas: They could not earn revenue through their extensive extortion networks, which required a powerful presence, and therefore their ability to intimidate and commit violence against the local population of the Federally Administered Tribal Areas diminished. All these direct and second-order effects of improved targeting yielded successful counter-terrorism results and also increased the acceptability and favorability of drone strikes among the local population of Pakistan’s tribal areas.
Can “Over-the-Horizon” Deliver?
After withdrawing from Afghanistan, the United States left behind non-state militant groups pitted against each other, fighting over turf and the resources of a war-torn state. Although a U.S. intelligence assessment, leaked to the New York Times, concluded that al-Qaeda does not have the ability to launch attacks from Afghanistan against America’s homeland, the Taliban cannot be trusted to ensure their territory is not used to plan and conduct terrorist attacks. In such a volatile environment and with no ground presence, the United States seeks to pursue an intelligence-forward, technology-driven strategy, dominated by unmanned systems, to ensure those groups cannot plan new attacks.
This “over-the-horizon” strategy is predicated on good, accurate human and signals intelligence. Over the last 20 years, the United States has established a strong, reliable intelligence gathering and analysis infrastructure across Central and South Asia. The U.S. withdrawal from Afghanistan will certainly result in diminished intelligence capabilities. However, according to the coordinator of the United Nations Analytical Support and Sanctions Monitoring Team that tracks non-state militant groups, “very rich sources” still remain in Afghanistan and the region. The United Kingdom, the United Nations through member nations, and a range of neighboring countries retain strong intelligence capabilities in Afghanistan. Although the U.S. departure resulted in a material reduction of direct intelligence collection, Washington maintains access to credible sources which can be triangulated, particularly with signals intelligence. Al-Zawahiri’s assassination validates this network. The al-Qaeda chief was communicating more freely after the Taliban takeover of Kabul, making him more vulnerable to electronic interception. This, possibly augmented by non-U.S. human intelligence networks and behavioral pattern-of-life analysis, allowed for an accurate identification.
Can “over-the-horizon” deliver? Yes. But is it sustainable? That is a more complicated question.
There are a lot of necessary factors: credible intelligence networks, cooperating governments, multilateral institutions with access, reliable unmanned combat and intelligence drones, surveillance and reconnaissance infrastructure, and targeting practices that adhere to the near-certainty standard. The United States managed to have all these in place during its drone operations in Pakistan from 2013 onwards. But it is a challenging task. Any variation could have adverse consequences and result in loss of life, like the tragically botched drone attack around the same time last year that killed 10 civilians, including seven children.
Unmanned warfare is no panacea. A counterterrorism strategy relying on unmanned assets is a short- to medium-term policy solution at best, and should be complemented with more long-term, non-military strategies. Yet as both the al-Zawihiri strike and my research in Pakistan show, drones have an important — and still sometimes underappreciated — role to play.
Dr. Neha Ansari is a postdoctoral fellow at the Center for Strategic Studies, Fletcher School of Law and Diplomacy. She studies emerging military technologies and their impact on warfare.