In reality, these “terrorists” included a civilian woman and a child.
AFRICOM, however, told the world a different story. “We assess no civilians were killed in this airstrike,” the command claimed, a regular feature of press releases following U.S. attacks in Somalia. A year later, AFRICOM publicly admitted those two, but only those two, civilian deaths. “This is our first confirmed civilian casualty incident in Somalia,” AFRICOM’s chief spokesperson, John Manley, told The Intercept in the aftermath of the reversal.
AFRICOM contends that hundreds of airstrikes and commando missions in the past 10 years have killed or injured only two civilians in Somalia. This flies in the face of scores of local accounts as well as investigations by international journalists and human rights organizations, including a recent report by Amnesty International. And The Intercept has obtained an AFRICOM document, through the Freedom of Information Act, that shows the command itself has long been aware of multiple attacks that left civilians dead or wounded following operations by U.S. or allied forces.
The document was produced by a team of AFRICOM personnel that conducts credibility assessments of civilian casualty, or CIVCAS, allegations. It contains brief synopses of alleged CIVCAS incidents, drawing on press coverage, reports from nongovernmental organizations, and self-reports by U.S. military personnel. Though undated, it lists incidents in 2018, so it appears to have been drawn up at the end of 2018 or early 2019. The document also presents AFRICOM’s own findings, including its determination of whether the allegation is “credible” or “not credible.” The document contains little detailed information about the reasons for those determinations, however, and there are no indications that AFRICOM conducts any shoe-leather investigations or interviews witnesses or family members of the victims of its attacks.
In 2018, according to the document, AFRICOM tracked 14 CIVCAS allegations in Somalia, two of which the command deemed “credible.” These incidents and AFRICOM’s credibility determinations — unlike the civilian deaths resulting from the U.S. attack on April 1, 2018 — have never been publicly disclosed until now.
The document, along with remarks from a former commander of U.S. Special Operations forces in Africa who spoke with The Intercept, suggests that AFRICOM may be classifying all military-aged males killed in airstrikes, including civilians, as combatants. (This has long been standard operating procedure in Afghanistan, suggesting that targeting protocols employed by U.S. Central Command have migrated to AFRICOM.) Experts say that even allegations deemed “not credible” by the command warrant further examination; yet there is no evidence that AFRICOM has conducted genuine investigations of CIVCAS allegations. AFRICOM is likely violating the laws of war when it comes to targeting, the experts say, and its claims about civilian casualties should, themselves, be deemed “not credible.”
“AFRICOM says it conducted 110 air strikes that killed 800 terrorists,” said Daphne Eviatar, director of the Security with Human Rights program at Amnesty International USA. “It’s just not plausible that all of the people killed were actually enemy armed forces, and that none were civilians.”
Who Is the U.S. Killing?
Even internal CIVCAS reports that are deemed “not credible” raise serious questions about who the U.S. is killing in its attacks. One such instance was “an airstrike targeting al-Shabab militants approximately 74 miles northwest of Mogadishu, Somalia, on August 2, killing four (4) terrorists,” according to a AFRICOM press release issued a day after the attack, on August 3, 2018. As usual, AFRICOM publicly concluded “no civilians were injured or killed in this airstrike.” Its internal assessment, while deeming CIVCAS allegations surrounding the strike as “non-credible,” raises disturbing questions about AFRICOM’s process for determining civilian casualties.
Amnesty International conducted an extensive investigation of the same incident, interviewing seven people — including witnesses of the attack, those who saw the corpses, and some who knew the victims — and concluded that this strike killed three civilians. They were a local businessman, 35-year-old Abdisamad Sheikh Issack Mohamed, and two well-diggers, 58-year-old Adan Hassan Yarow and 58-year-old Abdi Nurow Adan, both of whom were married and each left behind nine children. Some witnesses interviewed by Amnesty suggested that a lone member of al-Shabab, identified by two people as “Malable,” was the fourth individual killed and the likely target of the strike. A senior local official also confirmed to Amnesty that three civilians were killed in the attack. “I don’t know why they were hit, but maybe it was a mistake,” the local official said. “The U.S. are making a lot of mistakes in this region.”
In a written response to Amnesty International, Africa Command dismissed the results of Amnesty’s investigation of the August 2, 2018, attack but did not address the ample evidence they provided. Instead, AFRICOM referred only to a report of the incident on somalimemo.net, a local Somali media outlet deemed by the command to be sympathetic to al-Shabab. There is no indication that AFRICOM’s credibility assessment is based on a genuine inquiry involving interviews and the gathering of evidence from all feasible channels. “As far as we can tell, AFRICOM doesn’t do any on the ground investigations, and none of the 150 people we interviewed had ever spoken to a government official, Somali or American, about these attacks,” Brian Castner, Amnesty International’s senior crisis adviser on arms and military operations, previously told The Intercept.
The AFRICOM document tells a different story from the command’s press release and raises the specter of civilian casualties, even in cases in which allegations were deemed “not credible.” According to the document, only “two of the four adult males in the vehicle were assessed as armed.” The document also notes that “several circumstantial factors indicate that the three adult males in the vehicle were associated with al Shabaab,” without explaining what “circumstantial factors” led to the determination, what an association “with al Shabaab” means, and why one of the deceased, apparently not even “associated with al Shabaab,” was nonetheless deemed a terrorist and not classified as a civilian.
“The deeply troubling reality is that military claims of ‘not credible’ civilian casualty reports are themselves not credible,” says Hina Shamsi, the director of the ACLU National Security Project.
Relaxed Targeting Standards
One of the reasons civilians may be killed by U.S. airstrikes and counted as combatants, according to experts, could be relaxed targeting standards.
In March 2017, President Donald Trump reportedly designated parts of Somalia as “areas of active hostilities,” meaning the lifting of Obama-era rules requiring that there be near certainty that noncombatants will not be injured or killed. “The burden of proof as to who could be targeted and for what reason changed dramatically,” said retired Brig. Gen. Donald Bolduc, who headed Special Operations Command Africa, or SOCAFRICA, from April 2015 to June 2017. Bolduc added that the change led AFRICOM to conduct airstrikes that previously would not have been carried out.
The U.S. government has little to say about these changes. “We do not comment on specific details of classified counterterrorism operations,” said a National Security Council spokesperson, who would only offer comment on the condition of anonymity, when asked to respond to new reports that Trump may have declared all of Somalia an “area of active hostilities.”
What is abundantly clear is that the number of U.S. airstrikes in Somalia have risen markedly under the Trump administration, jumping from 14 under President Barack Obama in 2016 to 47 last year. The U.S. is on track to conduct at least 140 airstrikes in Somalia in 2019 if it maintains its current pace, according to Amnesty International.
“Many senior leaders have been removed from the battlefield, forcing [al-Shabab] to replace them with less effective leaders,” Gen. Stephen Townsend, the Trump administration’s pick to be the next head of AFRICOM, explained in written answers to questions from the Senate Armed Services Committee in April. “Airstrikes have also put pressure on the network responsible for car bomb attacks in Somalia’s capital, resulting in a one-third decrease in such attacks in 2018 … the first reduction in years.”
Questions have long swirled around such attacks and who, exactly, is dying in U.S. airstrikes. New America, a Washington, D.C.-based think tank, concludes that as many as 53 civilians have been killed in U.S. attacks in Somalia since 2003. The London-based Bureau of Investigative Journalism puts the toll at as many as 58 killed, although experts believe these may be low estimates. Amnesty International’s recent investigation of just five of more than 80 alleged U.S. airstrikes in Somalia between April 2017 and December 2018 — including the August 2, 2018, airstrike — found credible evidence that 14 civilians, including children, were killed.
According to Bolduc, all military-aged males are considered legitimate targets if they are observed with suspected al-Shabab members in locations that the U.S. classifies as supportive of the terrorist group. This is counterproductive, he says, because airstrikes don’t affect ideology and may end up breeding greater instability. “You just can’t go in there and kill everything that moves,” said the former SOCAFRICA commander. “I don’t have anything against HVT [high-value target] hunting, but we can’t continue to destroy everything in our path in the process of trying to secure U.S. national objectives. Because, at the end of the day, we’ve done nothing to change the fundamental security and stability of the environment.”
AFRICOM’s internal assessment of the August 2, 2018, strike, says Eviatar, “seems to confirm exactly what the Department of Defense has vehemently denied, which is that military age males in the vicinity of known al Shabaab leaders in areas believed to be sympathetic to al Shabaab are considered terrorists and therefore legitimate targets.” The document, she notes, confirms Bolduc’s assertions about classifying all military-aged males killed in airstrikes as terrorists and raises serious questions about AFRICOM’s targeting efficacy. “Obviously it’s a difficult situation when you’re fighting a group that doesn’t wear uniforms, but we supposedly have enough intelligence that we’re targeting people, so we should have enough intelligence about who we’re targeting,” Eviatar said. “Under the law, if there is a doubt as to whether someone is a civilian or a combatant, then you’re supposed to assume they are a civilian.”
AFRICOM disputed Bolduc’s and Amnesty’s characterization of its targeting standards but refused to offer any clarity about its methods. “I cannot provide additional detail due to operational security reasons,” Manley told The Intercept. Eviatar pointed out, however, the weakness of Manley’s “operational security” excuse. “These battles are over by weeks, months, even years,” she noted. “It’s just not credible that there is such sensitive information about why you chose a particular target that it can’t be released after all this time.”
Blaming Human Shields
Airstrikes are only one facet of U.S. military activities in Somalia. American commandos have also long accompanied local troops in ground operations. Such efforts include 127e, or “127 echo,” programs, named for the budgetary authority that allows U.S. Special Operations forces to use various foreign military units as proxies in counterterrorism missions. In recent years, these have been carried out in Somalia by local forces — including Kenyans and Somalis — under the code names Exile Hunter, Kodiak Hunter, Mongoose Hunter, Paladin Hunter, and Ultimate Hunter. Other operations, like Jupiter Garret, are run by Joint Special Operations Command, the secretive organization that controls the Navy’s SEAL Team 6 and Army’s Delta Force, and specifically hunt high-value targets.
These and other activities involving Americans and local forces have led to publicly acknowledged U.S. combat deaths. They have also resulted in allegations of killings of civilians. AFRICOM’s CIVCAS file shows that the command has its own evidence of credible allegations of previously unreported civilian casualties during these types of missions.
On January 19, 2018, for example, American-backed local forces attacked an Islamic school near Jameeco Jilyaile in the Middle Shebelle region, killing five children and a teacher, according to somalimemo.net. The Associated Press and CNN reported on the incident but characterized it as a joint raid by Somali and U.S. commandos to free child soldiers from an al-Shabab camp.
The AFRICOM CIVCAS document contains almost no details about the incident but indicates that the command reviewed the CIVCAS allegation and deemed it “credible.” When the internal inquiry was closed, about a month after the attack, AFRICOM’s internal CIVCAS evaluation acknowledged that the mission had resulted in civilian casualties. These were, however, blamed on al-Shabab employing “human shields.” AFRICOM never publicly acknowledged these deaths and paid out no compensation to the families of victims or survivors of the attack.
Join Our Newsletter
Original reporting. Fearless journalism. Delivered to you.
“Obviously, using human shields is a violation of the laws of war,” said Amnesty International’s Eviatar. “And if the U.S. determines that a target has such military significance that it’s worth killing a certain number of civilians, they’re allowed to do that under the laws of war. But it’s always difficult to assess because the military doesn’t give us the information about how they reach that conclusion. And just because a militant group is using human shields doesn’t excuse the U.S. government from treating them as civilians and protecting them as civilians.”
That isn’t the only allegation deemed “credible” in AFRICOM’s CIVCAS files. About a month later, U.S. troops were operating alongside Kenyan counterparts near Sungunni, Somalia. A firefight erupted between the Kenyan Defense Forces and al-Shabab militants, according to the CIVCAS documents, and a 20-year-old civilian woman was wounded by small arms fire. U.S. forces claimed that they never fired their weapons and that it was impossible to know who shot the young woman in the arm. After treating the 20-year-old, U.S. troops reported the incident to superiors, leading AFRICOM to deem a second CIVCAS report in less than a month “credible.” Neither the casualty nor the mission was ever publicly disclosed, and no compensation was paid to the injured woman.
Last year, Gen. Joseph Dunford, the chairman of the Joint Chiefs of Staff, commissioned a study of civilian casualties resulting from U.S. operations in the Central Command and AFRICOM areas of operation from 2015 to 2017. It specifically called for the military to “clarify guidance and doctrine to address the increased risk of CIVCAS when U.S. forces operate by, with, and through partner forces whose interests may not align with those of the United States.”
In written responses to questions from Sen. Elizabeth Warren about implementation of this recommendation, posed in April and obtained by The Intercept via the Freedom of Information Act, the incoming AFRICOM commander, Gen. Stephen Townsend, replied that allies receive training in the laws of war and “instruction on methods to prevent civilian casualties, including marksmanship, fire discipline, and target discrimination.” Experts, however, question whether this is enough.
“If the Kenyan forces who caused the casualties were operating under 127e authority, so that U.S. forces were directing their actions, then the U.S. could be responsible for the outcome and should be investigating those incidents,” says Amnesty’s Eviatar.
“It Strains Credulity”
Despite internal documents that detail “credible” reports of incidents that left civilians dead or wounded and “not credible” allegations that raise serious questions about who AFRICOM is killing in its airstrikes, the command continues to accept responsibility for only two civilian casualties after a decade of attacks.
“Speaking solely for U.S. Africa Command, to date we have not had any other confirmed reports of any other civilian casualties,” spokesman John Manley told The Intercept. “To date, only one civilian casualty assessment has been deemed credible, which is the April 1, 2018 airstrike.” This contention was echoed in a recently released, congressionally mandated Department of Defense report.
“That’s just not how war works, especially not air wars,” counters Eviatar. “We know that al Shabaab members are integrated into communities in Somalia, they don’t just occupy isolated military bases, so it strains credulity to suggest that US drones and manned aircraft dropping bombs in areas where civilians live and farm and congregate only killed al Shabaab fighters.”
“The military’s evidentiary standard for assessing credibility for civilian casualty reporting is so high that it can result in misleading public claims,” said Hina Shamsi, of the ACLU. “Its opaque investigative practices beg the question whether they can truly be effective. Most problematically, in incident after tragic incident, the basic issues are whether the military has accurate and reliable intelligence before it lethally strikes, and if it is complying with law that safeguards against civilian harm. Too often, the answers appear to be no.”
Eviatar argues that it’s time to stop taking AFRICOM at its word. “They’re just saying ‘trust us,’ and we can’t trust them because they’ve already made outlandish claims, like ‘Of the 800 people who were killed in 110 air strikes, none were civilians,’” she said. “There’s just no reason for human rights groups or anyone else to believe them. They don’t have to demonstrate that what they’re doing is lawful, and yet they can go out and kill people. It’s infuriating because what they’re really saying, effectively, is that they’re above the law.”
This article appeared earlier in theintercept