In March 2019, the U.S. military launched airstrikes against the last ISIS hold-out in Baghuz, Syria. Despite official Department of Defense (DOD) reporting that only four civilians were killed in the strikes, new reporting from the New York Times indicates that it was one of the largest civilian casualty incidents in the war against ISIS. In response to the Times investigation, U.S. Central Command (CENTCOM) publicly acknowledged for the first time that 80 people were killed, and that it was unable to determine conclusively the civilian or combatant status of 60 individuals.

We provide a list of specific questions for members of Congress, reporters, and investigators to ask about the strike. 

Following the investigation into the tragic Aug. 29 drone attack in Kabul, the recently revealed Baghuz strikes raise profound questions about DOD’s policies and practices for conducting strikes and investigating errant attacks. In particular, the strikes raise questions concerning target identification and verification practices, proportionality assessments, and post-strike investigations into civilian casualties. The Baghuz strikes also raise serious questions about investigations and accountability in cases where such strikes could amount to a war crime. DOD must address these questions if it is to restore public confidence in its operations.

I. Pre-Strike Questions

General Framework and Procedures

1. Why has the DOD report concerning the investigation into the Baghuz strikes not been made public? Will DOD commit to releasing an unclassified version of the report or its findings?

2. What was the legal authority under which the strikes were conducted? 

3. Were senior Pentagon officials, others in the executive branch, or members of Congress informed of the Baghuz strikes? If so, when were they briefed? Did any of those officials request additional investigation or review of the strikes?

4. As the International Committee of the Red Cross explains, the presumption of civilian status is a part of the binding laws of war. Does DOD still hold the highly anomalous view (compared to U.S. allies and globally) that this rule is not part of the binding laws of war? Does the U.S. government as a whole agree with DOD’s position?

5. Why were U.S. special forces reportedly unaware of the location of coalition assets in theater that might have reduced the risk of civilian harm? Does DOD have a list of such assets and, if so, why wasn’t this list shared with operators on the ground? 


6. What are the criteria for determining whether a strike may be considered “unit self-defense”? At what level are such decisions approved?

7. What are the criteria for determining whether a strike may be considered “collective self-defense” of a partner force? At what level are such decisions approved?  

8. In cases of collective self-defense, what procedures are in place to verify that partners are indeed facing an imminent threat? 

9. To what extent does DOD rely on partner information when making determinations about whether to authorize a strike, and how much does this information affect pre-strike proportionality and civilian casualty assessments? When such information conflicts with information from aerial surveillance or other available sources, how are these different information streams reconciled prior to launching an attack?

10. Are there any procedures in place to ensure that strike logs accurately reflect information that is seen on aerial surveillance feeds? 

Civilian Casualty Assessments 

11. What pre-strike procedures did DOD follow to assess whether there were civilians in the area and how long did this assessment take? 

12. Before taking the strike, what did DOD estimate would be the total number of civilians killed or injured? 

13. What level of certainty did DOD have regarding the estimated number of civilians killed or injured?

14. What did DOD consider would have been an acceptable level of civilian casualties before authorizing the strike?

15. Did DOD apply the standard, which is codified in the Geneva Conventions (Additional Protocol I), that in case of doubt, people and objects should be presumed to be civilians?

Target Identification and Engagement 

16. According to DOD, civilians were not observed on the standard definition camera used by the task force (TF) that conducted the strike, but “a coalition UAV operator”  had a high-definition camera present at the time of the strike that did show the presence of civilians. Why was the TF not aware of the presence of this other asset? Has this happened previously in Syria, Iraq, or other theaters and, if so, what remediation actions were supposed to have been put in place to prevent it from happening again?

17. DOD determined that 16 people killed were ISIS fighters and four were civilians. On what basis did they make those determinations and with what level of confidence? 

18. Why was DOD unable to determine the combatant or civilian status of the other 60 people? What efforts has it undertaken to assess the status of these individuals over the past 2.5 years? What have been complicating factors preventing a high confidence assessment? How many of the 60 were women? Children?

19. Does DOD consider anyone who is armed to be a combatant, regardless of whether they are directly participating in hostilities?

20. How did DOD ensure that the strikes were conducted in accordance with the principles of necessity, distinction, and proportionality? 

21. The principle of proportionality requires that the expected harm is not excessive in relation to the anticipated concrete and direct military advantage, and that feasible precautions are taken in planning and conducting attacks to reduce the risk of harm to civilians. Were feasible precautions taken to reduce the risk of civilian harm and, if so, what were the precautions?

22. What was the anticipated military advantage and how was that determined?

23. Would DOD still have considered the strikes to be proportional if it had determined in advance that there were more than 60 civilians present in the target area? What factors does DOD take into consideration when making determinations of proportionality? 

24. What are the criteria that leads DOD to determine that a woman or a child is a combatant? DOD notes that “some women and children, whether through indoctrination or choice, decided to take up arms in the battle and as such could not be strictly classified as civilians.” What about those who are forced to participate, or at least to carry a weapon — would that change any of DOD’s assessments? 

25. DOD notes that it relied on its local partners, in part, to assess the presence of civilians. Is this a common practice, and how much confidence does DOD place on these assessments compared to its own pre-strike intelligence?

26. Can DOD provide more information on how the 80 individuals targeted in this strike posed a threat to Syrian Defense Forces (SDF)? DOD noted that 16 active ISIS fighters from that group were engaged in attacks on SDF positions. What were the other 64 doing at that time?

27. CENTCOM stated that the strikes using 500- and 2000-pound bombs “were proportional due to the unavailability of smaller ordinance at the time.” That is not a recognizable version of the standard of proportionality. Can DOD clarify how proportionality was applied to this strike?


I. Post-Strike Questions

Initial Assessments 

28. What observations of the strike location did DOD or other agencies carry out post-strike? For how long were observations of the location carried out? What methods were used, other than aerial surveillance? Was new surveillance conducted or only the surveillance available in the immediate period following the strike?

29. What training is given to operators conducting post-strike assessments and do these individuals have access to the full range of classified information needed to accurately conduct such assessments? 

30. To what extent did DOD rely on information provided by the SDF or other local partners in reaching its assessment of civilian casualties?

31. Why did DOD fail to report the assessed four civilian casualties in its 2019 or 2020 civilian casualty reports to Congress? How does DOD intend to report these numbers now? For the 60 for which DOD cannot conclusively characterize their status, would they be presented as a range of potential civilian casualties? What is DOD’s standard practice for publicly accounting for deaths for which it cannot conclusively assess their status? 

32. When reporting civilian casualties, the Office of Director of National Intelligence has stated that when it cannot be determined if an individual was a combatant, they must be assessed as a civilian: “The assessed range of non-combatant deaths includes deaths for which there is an insufficient basis for assessing that the deceased is a combatant.”

Why was this standard not followed in this case?

33. Which allies under the Global Coalition to Defeat ISIS knew about this strike? Did any of them call for additional investigations or raise concerns about the legality of the strike?

34. Has DOD made any attempt to provide solatia or ex gratia payments to the families of the civilians killed at Baghuz?

35. Under the civilian casualties Executive Order 13732, DOD is required to consider “relevant and credible information from all available sources, such as other agencies, partner governments, and nongovernmental organizations” in assessing civilian casualties. Did DOD solicit this wide range of information? What information did non-governmental organizations (NGOs)  including media, local third parties like Raqqa is Being Slaughtered Silently, and international human rights provide on the strikes?


36. When conducting the initial investigation, what steps did DOD take to respond to internal concerns raised by DOD officials and external allegations from NGOs that the strikes killed dozens of women and children? 

37. Did DOD or coalition partners bulldoze the target area immediately following the strikes? What was the reason for this? If not, does DOD know who took such reported actions?

38. Is it the case that the same operators who authorized the strikes were responsible for the initial post-strike assessment? How can post-strike assessments be accurate and objective if operators are responsible for grading their own behavior?  

39. What actions did the Senate Armed Services Committee take in response to the information that Colonel Korsak shared regarding potential war crimes? Why has he not received a response from the committee? 

40. Is it the case that the Air Force Special Investigations office only investigates civilian casualty incidents where there is a “potential for high media attention, concern with outcry from local community/government, [and/or] concern sensitive images may get out”? Why was that criteria cited in a response Col. Korak reportedly received from the Office of Special Investigations?

41. Why was the assessment that a violation of the law of armed conflict may have taken place removed from the final draft of the civilian casualty report? 

42. Why were all mentions of the Baghuz strikes cut from the inspector general report on civilian casualties that was formally released this spring to select members of Congress and the military? 

43. Why did the deputy inspector general refuse to sign a memo in January 2020 that would have formally alerted authorities and opened a criminal investigation? 

44. Why is Operation Inherent Resolve still publicly listed as an “open” case on DOD’s website 2.5 years after the official civilian casualty report determined that only four civilians were killed? 

45. Is DOD currently drafting a report evaluating CENTCOM’s adherence to the laws of war? Will DOD commit to releasing an unclassified version of the report or its findings?

46. In light of the systemic problems that recent revelations have raised, including the Aug. 29 Kabul strike and the March 2019 Baghuz strikes, will DOD commit to releasing unclassified versions of any reports in which similar allegations have been made that U.S. special forces hit targets knowing civilians would be killed? 

Accountability and Future Operations

47. DOD notes that “the investigating officer determined that no disciplinary actions were warranted.” Did the chain of command determine that other action was appropriate? Short of criminal action, did any personnel involved in the strike receive administrative punishment, decertification, or other remedies that the Department previously suggested were available?

48. What steps has DOD taken to ensure that similar civilian casualty incidents do not occur in the future? Has it changed any targeting methodologies, pre-strike procedures, or post-strike assessment practices? 

Image: Smoke billows after shelling on the Islamic State group’s last holdout of Baghouz, in the eastern Syrian Deir Ezzor province on March 3, 2019. (Photo by DELIL SOULEIMAN/AFP via Getty Images)