Deputy Secretary of Defense Patrick Shanahan had high hopes when he initiated a high profile effort to accelerate the adoption of a department-wide cloud computing system” in September 2017. Cloud computing systems would be “critical to maintaining our military’s technological advantage,” he said.
Yet despite all the urgency and the high-level support, three and a half years later, the solution—the Joint Enterprise Defense Infrastructure (JEDI) cloud program—remains mired in bureaucracy. Our warfighters still cannot reap the benefits of enterprise cloud computing.
With the soon-to-be released Joint Warfighting Concept of All-Domain Operations calling for unprecedented levels of collaboration between services, the need could not be more urgent.
The delay in getting JEDI capability illustrates just how far the Department of Defense (DoD) needs to go if it hopes to successfully compete in great power competition.
In 2017, the Pentagon was already far behind the commercial sector in adopting cloud computing—though not for lack of trying. DoD had over 500 disparate cloud initiatives underway and maintained hundreds of its own data centers, each with its own versions of software, hardware, and infrastructure. These balkanized efforts multiplied security risks, frustrated collaboration, and carried huge costs. Lack of an all-classification enterprise cloud computing capability denied the military an ability to quickly connect its force, from infantry squads to defense agencies.
After crafting their acquisition strategy and adjudicating over 1,500 comments, the DoD released the JEDI Request for Proposal on July 26, 2018. At this point, the effort was still moving with some alacrity. Multiple companies submitted bids.
When the Pentagon decided to make a single award to only one provider, it sparked great resistance from industry and trade groups, which lobbied Congress for relief. Industry saw JEDI as a potential gold mine worth up to $10 billion. Congress started asking questions and asking for reports.
DoD, however, remained steadfast in its pursuit of a single award, explaining that the JEDI contract was a "pathfinder" and that to reduce complexity in this initial effort, a single award was necessary.
Absent significant intervention from Congress, industry went in another direction. Both Oracle America and IBM filed protests even before an award was made. The GAO subsequently denied both. Following GAO’s denial of its protest, Oracle then filed a lawsuit with the U.S. Court of Federal Claims in December 2018. The court ruled against Oracle in July 2019.
After earlier down-selecting to two companies: Amazon Web Services and Microsoft, the Pentagon announced on October 25, 2019, that it had awarded the JEDI contract to Microsoft.
The following month, Amazon filed a protest with the Court of Federal Claims, and in February 2020, the court issued an injunction prohibiting any work from beginning on JEDI.
Amazon’s protest was based on two main claims: that “egregious errors” were made in evaluating the proposals and that then-President Trump exercised undue influence in the selection process.
The matter remains unresolved, with no decision as yet from the Court of Federal Claims, which seems unaware of the importance of this capability to the U.S. military.
Without opining on the merits of these protests, one can draw the obvious conclusion that is something is deeply wrong with a system that takes three and a half years to craft, award, and initiate work on a services contract that enjoyed the personal priority of the Deputy Secretary of Defense.
Note: We are not talking about designing and manufacturing a new stealth bomber or an aircraft carrier. This is about securing the services of a commercial cloud services company to provide well-understood support services—just like what they provide thousands of other entities.
There is blame enough for everyone.
The Pentagon did itself no favors by not making an early, compelling case for the need for a JEDI capability and why a single award was necessary. Then, after acknowledging some flaws in their award process, DoD found itself forced to “re-award” the contract to Microsoft in September 2020. Better performance must be expected.
Companies are obligated to look out for their stockholders' best interests and should be expected to make a protest when they feel legitimately wronged by the process. But at some point, a line is crossed when repeated protests start to feel more like obstructionism than relief from injustice.
The Court of Federal Claims has had this case since November 2019 and appears to share none of the Pentagon’s urgency to acquire this capability. Yes, a wholly independent judiciary answers to no one, and this is how we want it. But there must be a means to obtain an accelerated judgment in the case of a national security priority.
While Congress didn’t intervene to stop the JEDI effort, they also didn’t help it by requiring multiple reports, strategies, and policies. Each took time to complete and diverted time and attention away from the main effort of acquiring the necessary capability.
The Pentagon has become so discouraged that officials now say, if the court doesn’t produce a decisive ruling soon, they might abandon the “urgent and unmet” JEDI cloud effort completely. That would be a misfortune that would be borne by our service members, not the denizens of Washington, D.C.
All these delays might be acceptable if the clock wasn’t running. But it is. The United States is in a life-or-death competition with other countries whose authoritarian systems are totally repugnant but do not suffer similar delays as what has befallen JEDI. Every second that ticks off the clock represents time that our military lacks the capability needed to fight successfully on the future's information-centric battlefield. America must do better.
This piece originally appeared in RealClear Defense