The Army 15-6 Investigation into the Airstrike

Within weeks of the airstrike the US Army had looked into it, written it up and closed the file. That inquiry is the only official examination of 12 July 2007 that exists, and almost every claim that the matter was investigated rests on it. It is worth knowing exactly what it was.

What an AR 15-6 actually is

An AR 15-6 is an administrative investigation named after the Army regulation that governs it. A commander appoints an officer to establish the facts of an incident and report back. It is used for all sorts of things: accidents, losses of equipment, deaths, complaints.

Three features that decide what it can find

Three features matter here.

Why that is not the same as worthless

• It is internal. The Army investigates itself, with an officer from inside the same structure.
• It is not criminal. A 15-6 cannot charge anyone. At most it can recommend that someone else consider charges.
• It serves a commander. Its purpose is to let a commander decide what to do, not to satisfy the public or the families.

None of that makes it worthless. It makes it a specific tool with a known shape, and reading it as though it were a court judgment is the most common mistake made about this case.

What it concluded

The investigation found that the aircrew acted in accordance with the rules of engagement and the law of armed conflict, given what they believed they were seeing. It accepted that the crew had identified what they took to be armed men, including at least one with a rocket-propelled grenade launcher, in an area where American ground troops were in contact.

It treated the deaths of the two Reuters staff as a tragic consequence of that identification rather than as a failure of it. No disciplinary action followed. The file was closed.

The questions it was not built to ask

The gap between the first row and the last is the whole problem. An inquiry that asks whether orders were followed will usually find that orders were followed, and that finding then gets reported as though the incident had been cleared in a wider sense. Following the rules and being in the right are different claims.

How anyone outside the Army ever read it

Not by publication. The report reached the public through freedom-of-information releases, in the slow, partial way such material usually does.

That timing matters. Reuters was asking for the footage and for answers from July 2007 onward and was refused for nearly three years. A closed internal inquiry, plus a refusal to release the recording it was based on, is precisely the combination that made the 2010 leak feel to many people like the only way the material was ever going to surface.

What the investigation does get right

It is easy to dismiss an internal inquiry, and worth resisting. The 15-6 established the sequence of the morning in detail, confirmed the units involved, and did not dispute that two Reuters staff were killed. A great deal of what this archive can state with confidence about the order of events rests on it.

Where its reasoning holds

It is also, on the first burst, a reasonable piece of reasoning. Given a crew looking at a grey, magnified image of men with something long on a shoulder, near troops taking fire, the conclusion that they believed they had identified a threat is not a stretch.

Where the investigation stopped, and never restarted

The Army closed the file in 2007 and it has never been reopened. When the footage became public in 2010 there were calls for an independent inquiry from news organisations and human rights groups. None was held.

Where that leaves the record today

So the position today is this. The only examination of these deaths was carried out by the institution responsible for them, within weeks, under a process that could not have brought a charge even if it had wanted to, and its conclusion has stood for nearly two decades without any outside body testing it. That is not the same as the matter being settled, and it is the reason the accountability question is still the loudest one in this archive.