The Army ordered the destruction of a report that praised the performance of an off-the-shelf software program that finds buried explosives in Afghanistan and replaced it with a revised, less-favorable assessment, according to internal Pentagon documents.
The unusual action came amid a battle inside the Army. It pits those who want the service to send more of the software platform, called Palantir, to the Afghanistan War against those who favor the Army’s own developed intelligence network, the Distributed Common Ground System (DCGS).
“We are trying to solve some very hard problems that pose life or death issues for the soldiers,” the officer emailed to the Pentagon.
The documents obtained by The Washington Times show that Gen. Raymond T. Odierno, Army chief of staff, in February ordered the Army Test and Evaluation Command (ATEC) to judge the performance of Palantir.
The Times first reported last week that commanders in Afghanistan asked higher-ups for permission to buy Palantir, as they raved about its ability to pinpoint a major killer of American troops — buried homemade bombs, or improvised explosive devices (IEDs).
Rep. Duncan Hunter, California Republican and a member of the House Armed Services Committee, said Army procurement officials at the Pentagon were trying to protect the DCGS, which the service developed with private industry, and discourage use of Palantir produced by Palantir Technologies in Palo Alto, Calif.
“What I’m concerned about the most is the bureaucracy in the Pentagon is stopping the war fighter from getting the right gear at the right time,” Mr. Hunter, who served in Afghanistan and Iraq as a Marine officer, told The Times.
“We now know in the past that these reports we’ve been getting on a lot of different things may have been revised internally. That’s a pretty damning thought.”
Events leading to the destruction of the favorable report began after ATEC published the 50-page analysis in April. Labeled “for official use only,” the assessment praised Palantir and offered some criticism of DCGS as too slow to process data.
On June 29, Col. Joseph Martin, who heads Army Operational Test Command, signed a memo killing the April report and replacing it with one dated in May.
A senior enlisted man in Afghanistan subsequently signed a memo that ordered the destruction of the April 25 report.
“Please ensure that any and all copies of the 25 April report are destroyed and not distributed,” the memo said.
“Upon destruction of the previous report, please confirm to me in email that all copies of the original report dated 25 April has been destroyed. If possible, can you please provide me the names of the individuals that the 25 April report was sent to.”
An Army spokesman told The Times on Sunday that the newspaper did not have the full story, but that it was difficult to come up with an official response over the weekend.
An internal email from an Army official, who defended the revision, told colleagues that the first report was killed because it contained inaccuracies and that the action was not taken to protect the common ground system.
In the “findings” section, for example, the second report deletes a paragraph that told of how Palantir allows war fighters to simultaneously search multiple databases, including information from British intelligence sources.
“This capability enables analysts to rapidly execute necessary data mining and create products that are requested by units for operations and missions more efficiently,” the erased paragraph said.
The second report removed a recommendation that the Army install more Palantir services in Afghanistan. It also deleted a recommendation that the Army institutionalize Palantir by creating formal training classes for intelligence officers.
In the recommendation section, the first report contained criticisms of the Distributed Common Ground System. It quoted a supervisor as saying, “DCGS is overcomplicated, requires lengthy classroom instruction, and is an easily perishable skill set if not used constantly.”
That comment was removed from the recommendation section but kept in the last pages of the report grouped with many other comments.
Frustrated intel officer
The Times also obtained documents showing the active effort by Army headquarters to keep Palantir from the field. The 82nd Airborne Division, based in regional command south in Afghanistan, had to prod the Army bureaucracy for permission to buy Palantir via the Army’s Rapid Equipping Force in the winter.
At one point, a frustrated 82nd Airborne intelligence officer wrote to higher-ups, “We are trying to solve some very hard problems that pose life or death issues for the soldiers under this command, and [DCGS] is not making our job easier, while Palantir is giving us an intelligence edge. This is a pretty big redline for many of the units in the field, of which 82nd Airborne Division is certainly the most visible.”
The officer added, “The chain of command believes they need to have this capability in the fight and that it will save soldiers’ lives and limbs. Bottom line, there is a significant capability gap in DCGS … that Palantir greatly exceeds, and with extremely high stakes in a very violent environment, today we need the capability advantage that Palantir provides.”
The officer offered a devastating assessment of the Army’s preferred system:
“Bottom line from our perspective is that [DCGS] has continuously overpromised and failed to deliver on capability that will meet the needs of the warfighter.
“All the bullet points they can list on a slide sitting back in the Pentagon don’t change the reality on the ground that their system doesn’t do what they say it does, and is more of a frustration to deal with than a capability to leverage.
“We aren’t going to sit here and struggle with an ineffective intel system while we’re in the middle of a heavy fight taking casualties. Palantir actually works. When DCGS actually works, we’ll be ready to use it.”
He added: “If the crew of people I work with their combined IQ, ingenuity, and years of experience, can’t figure out how to make DCGS work in this fight, they need to fix their system.”
“If they would have gotten this software six months earlier or a year earlier, or be able to buy into it prior to deployment, who knows how many lives could have been saved?” he said.
“Who knows how much more successful they could have been at finding IEDs and finding the guys who were making the IEDs?”